Re: Pertec Interface Cable Length
On 8/12/19 8:16 AM, Nigel Johnson via cctalk wrote: > Another consideration with the TC02 is the small buffer. I don't know > what tape speed your drive runs at, but we lost a lot of sales to Dilog > because of buffer overflow on some of the faster CDC dirves. When we > came out with the TC03, it had a larger buffer to handle this. The Qualstar drive is a *slow* drive. 50 ips in 1600 PE mode and a whopping 12.5 IPS in 6250 GCR mode. Transfer rate in either is about 80KB/sec. Qualstar doesn't have a mechanical buffer (spring arms or vacuum column) like most drives; it's all handled by the reel servos. It's wonder that it works at all.I doubt that it will overwhelm a PDP11. It probably doesn't overwhelm an IBM PC XT. The SCSI version of the drive does have a 256KB buffer, but I'm uncertain what the Pertec interface model has. --Chuck
Re: Pertec Interface Cable Length
On 08/11/2019 08:00 PM, Douglas Taylor via cctech wrote: This is where the electrical engineer could help. How do you determine how long a cable the 74LS240 can drive? Well, there are several considerations. First, it takes some current to charge up the cable capacitance. More current charges the capacitance faster, but also creates faster edges which cause more crosstalk. Then, the data rate needs to be considered. Mag tape data rates are not that high. So, for 1600 BPI at 45 IPS, the data rate is 72 K bytes/second, or about 14 us per byte. Twisted-pair cable should have a little less capacitance, and it is supposed to reduce crosstalk, so should work better. The most serious problem is when many data lines switch at the same time, it may contaminate the clock pulses and cause bytes to be dropped or added. With the low data rates involved, proper delays to allow ringing to settle on the data lines and prevent short crosstalk pulses from affecting the clocks should make the system very tolerant of cable issues. But, maybe some engineers didn't really optimize their logic for these problems. Jon
Re: Pertec Interface Cable Length
On 8/12/19 8:11 AM, Douglas Taylor via cctech wrote: > The bad news is that the cable lengths must be short to use the Qualstar > 1260 with a PDP11, the good news is that I can lift and carry the tape > drive! For many of us in this hobby that it is extremely important. > > After looking at pictures of the 1260 on the internet I see that it was > designed to be used with a PC and the interface cable was 62 pins and > quite short. Someone mentioned earlier that it was a cheap tape drive > that didn't meet the Pertec standard and I'm finding out what exactly > that meant. > > It is nice to have a reel to reel tape drive and watch it work. If you needed to cobble something up suitable for long cable driving, you could work up the correct pertec driver interface to sit between the Qualstar interface board and the cable and install it in the 1260 case--there's plenty of room without the SCSI interface PCB. The big problem with the 1260 is that it doesn't move the tape fast enough for reliable operation at the 6250 GCR setting; operation at 1600 PE is just passable. But it's a drive that's portable and that is an advantage, especially to us older folk. Another possible option would be to replace the Qualstar LS240 drivers with TI 74BCT756 open-collector drivers (same pinout) with 64 ma drive capability. That probably would be the harder option, as it would entail removing the soldered-in LS240s. FWIW, Chuck
Re: Pertec Interface Cable Length
Another consideration with the TC02 is the small buffer. I don't know what tape speed your drive runs at, but we lost a lot of sales to Dilog because of buffer overflow on some of the faster CDC dirves. When we came out with the TC03, it had a larger buffer to handle this. cheers, Nigel Johnson On 12/08/2019 11:13, Jon Elson via cctech wrote: On 08/11/2019 08:00 PM, Douglas Taylor via cctech wrote: This is where the electrical engineer could help. How do you determine how long a cable the 74LS240 can drive? Well, there are several considerations. First, it takes some current to charge up the cable capacitance. More current charges the capacitance faster, but also creates faster edges which cause more crosstalk. Then, the data rate needs to be considered. Mag tape data rates are not that high. So, for 1600 BPI at 45 IPS, the data rate is 72 K bytes/second, or about 14 us per byte. Twisted-pair cable should have a little less capacitance, and it is supposed to reduce crosstalk, so should work better. The most serious problem is when many data lines switch at the same time, it may contaminate the clock pulses and cause bytes to be dropped or added. With the low data rates involved, proper delays to allow ringing to settle on the data lines and prevent short crosstalk pulses from affecting the clocks should make the system very tolerant of cable issues. But, maybe some engineers didn't really optimize their logic for these problems. Jon -- Nigel Johnson MSc., MIEEE VE3ID/G4AJQ/VA3MCU Amateur Radio, the origin of the open-source concept! You can reach me by voice on Skype: TILBURY2591 If time travel ever will be possible, it already is. Ask me again yesterday This e-mail is not and cannot, by its nature, be confidential. En route from me to you, it will pass across the public Internet, easily readable by any number of system administrators along the way. Nigel Johnson Please consider the environment when deciding if you really need to print this message
Re: Pertec Interface Cable Length
On 8/12/2019 1:25 AM, Chuck Guzis via cctech wrote: On 8/11/19 8:51 PM, Douglas Taylor via cctech wrote: The TC02 is an Emulex TS11 emulation for pertec interface tape drives. The J1 and J2 are sort of standard terminology, don't know why. Ah, the *Emulex* TC02. You had me going there--DEC also has a DECtape controller called the TC02. Looking at the TC02, there are 374s to latch data coming from the Qualstar and use the termination packs, but there are also 7438s driving the lines from the TC02 to the Qualstar. Those have no terminators. The TC02 reference manual says that you get run lines up to 30 feet long between the TC02 and formatter. My point is that the driver technology for the Qualstar (i.e. read data and status) is inappropriate for long cable runs. The spec calls for 48 ma OC drivers. --Chuck , The bad news is that the cable lengths must be short to use the Qualstar 1260 with a PDP11, the good news is that I can lift and carry the tape drive! For many of us in this hobby that it is extremely important. After looking at pictures of the 1260 on the internet I see that it was designed to be used with a PC and the interface cable was 62 pins and quite short. Someone mentioned earlier that it was a cheap tape drive that didn't meet the Pertec standard and I'm finding out what exactly that meant. It is nice to have a reel to reel tape drive and watch it work. Doug
Re: Pertec Interface Cable Length
On 8/11/19 8:51 PM, Douglas Taylor via cctech wrote: > > The TC02 is an Emulex TS11 emulation for pertec interface tape drives. > The J1 and J2 are sort of standard terminology, don't know why. Ah, the *Emulex* TC02. You had me going there--DEC also has a DECtape controller called the TC02. Looking at the TC02, there are 374s to latch data coming from the Qualstar and use the termination packs, but there are also 7438s driving the lines from the TC02 to the Qualstar. Those have no terminators. The TC02 reference manual says that you get run lines up to 30 feet long between the TC02 and formatter. My point is that the driver technology for the Qualstar (i.e. read data and status) is inappropriate for long cable runs. The spec calls for 48 ma OC drivers. --Chuck
Re: Pertec Interface Cable Length
On 8/11/19 9:11 AM, W2HX via cctech wrote: >> All of my Pertec tape drives (in the past) were located in > a differentrack from the actual computer so the cables were > always over 10' long and usually close to if not 20'. Never > had a problem. I stand by my commentary re Qualstar 1xxx drives. They're not up to the task of driving long cables. Who has one such drive with 10' cables operating correctly? Until someone comes up with actual experience with said drives, we're just guessing. I note that my 1260S does have the capability to be used as a Pertec interface drive. I suppose I could re-cable from one of my other drives and test it, but at this stage, I hardly see the point. --Chuck
Re: Pertec Interface Cable Length
On 8/11/19 6:00 PM, Douglas Taylor via cctech wrote: > I just looked at the TC02 and the Qualstar, there are termination > resistor packs on each. The Qualstar has a bunch of 74LS240 IC's near > the J1 and J2 pertec interface cables. The TC02 has a bunch of 74LS374 > chips near the J1 and J2 connectors. > > This is where the electrical engineer could help. How do you determine > how long a cable the 74LS240 can drive? TC02? The DECtape controller? Sorry, I must be dense; I'm not following. --Chuck
Re: Pertec Interface Cable Length
On 8/11/2019 10:44 PM, Chuck Guzis via cctech wrote: On 8/11/19 6:00 PM, Douglas Taylor via cctech wrote: I just looked at the TC02 and the Qualstar, there are termination resistor packs on each. The Qualstar has a bunch of 74LS240 IC's near the J1 and J2 pertec interface cables. The TC02 has a bunch of 74LS374 chips near the J1 and J2 connectors. This is where the electrical engineer could help. How do you determine how long a cable the 74LS240 can drive? TC02? The DECtape controller? Sorry, I must be dense; I'm not following. --Chuck The TC02 is an Emulex TS11 emulation for pertec interface tape drives. The J1 and J2 are sort of standard terminology, don't know why.
Re: Pertec Interface Cable Length
On Sun, Aug 11, 2019 at 7:44 PM Chuck Guzis via cctech wrote: > > On 8/11/19 6:00 PM, Douglas Taylor via cctech wrote: > > > I just looked at the TC02 and the Qualstar, there are termination > > resistor packs on each. The Qualstar has a bunch of 74LS240 IC's near > > the J1 and J2 pertec interface cables. The TC02 has a bunch of 74LS374 > > chips near the J1 and J2 connectors. > > > > This is where the electrical engineer could help. How do you determine > > how long a cable the 74LS240 can drive? > > TC02? The DECtape controller? > I assume TC02 here refers to an Emulex TC02, a Q-Bus Pertec interface tape controller which emulates a TS11 tape controller. http://www.bitsavers.org/pdf/emulex/TC0251002-G_TC02tech_Jul85.pdf
Re: Pertec Interface Cable Length
On 8/11/2019 7:01 PM, Jon Elson via cctech wrote: On 08/11/2019 11:11 AM, W2HX via cctech wrote: I seem to remember they were ribbon cables with each odd/even pair twisted which probably meant one active and one ground twisted together. Or differential pairs. No, both Pertec unformatted and Pertec formatted interfaces were TTL single-ended. Jon I just looked at the TC02 and the Qualstar, there are termination resistor packs on each. The Qualstar has a bunch of 74LS240 IC's near the J1 and J2 pertec interface cables. The TC02 has a bunch of 74LS374 chips near the J1 and J2 connectors. This is where the electrical engineer could help. How do you determine how long a cable the 74LS240 can drive? Doug
Re: Pertec Interface Cable Length
On 08/11/2019 11:11 AM, W2HX via cctech wrote: I seem to remember they were ribbon cables with each odd/even pair twisted which probably meant one active and one ground twisted together. Or differential pairs. No, both Pertec unformatted and Pertec formatted interfaces were TTL single-ended. Jon
Re: Pertec Interface Cable Length
>All of my Pertec tape drives (in the past) were located in a differentrack from the actual computer so the cables were always over 10' long and usually close to if not 20'. Never had a problem. That is because a) you had the proper termination in place and b) the signalling speed of the interface was slow enough to cope (you were probably within spec). > I seem to remember they were ribbon cables with each odd/even pair twisted which probably meant one active and one ground twisted together. Or differential pairs. 73 Eugene W2HX From: cctech on behalf of Bill Gunshannon via cctech Sent: Sunday, August 11, 2019 9:59 AM To: cct...@classiccmp.org Subject: Re: Pertec Interface Cable Length On 8/11/19 9:07 AM, W2HX via cctech wrote: >> Would folding the > excess cable up and covering with anti-static plastic help? > > unlikely. The typical problem is that the longer the cable, the higher the > capacitance of the transmission line. Therefore you get a lot of problems > with rising and trailing edges of the signals which can cause all kinds of > problems if not terminated correctly. I agree with the previous suggestion > to make sure all termination is in place. If you can also place a scope on > one of the lines (with high impedance probe) you can see what the edges look > like. All of my Pertec tape drives (in the past) were located in a differentrack from the actual computer so the cables were always over 10' long and usually close to if not 20'. Never had a problem. I seem to remember they were ribbon cables with each odd/even pair twisted which probably meant one active and one ground twisted together. > > From: cctech on behalf of Douglas Taylor via > cctech > Sent: Sunday, August 11, 2019 12:33 AM > To: Jon Elson; On-Topic Posts > Subject: Re: Pertec Interface Cable Length > > On 8/10/2019 1:56 PM, Jon Elson wrote: >> On 08/09/2019 11:05 PM, Douglas Taylor via cctech wrote: >>> I have a question about cable length - any electrical engineers in >>> the house? >>> >>> Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape >>> controller in a pdp-11/53. The interface is pertec with 2 50 pin >>> cables. >>> >>> When I use a pair of short flat ribbon cables, 18 and 30 inches each, >>> it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. >>> >>> It doesn't work when I use a pair of 5 foot long flat ribbon cables. >>> Are they too long? Do I need twisted pair type of cable? Is it >>> possibly a termination problem? >>> >> I have used cables about 20 feet long without trouble. The 2 50-pin >> cables is the Pertec formatted interface, which is really forgiving. >> Does you drive have terminators in both ends of the cable (both at the >> TC02 end and the drive end)? >> Now, I will mention that I have ONLY used twisted-pair ribbon cables >> with both flavors of interface, never straight ribbon cable. >> >> Jon >> >> > I haven't checked to see if there are terminators (Arnold the > Terminator) on either end. I did check the long cables for continuity > and found no problems. It may be an EMI problem. Would folding the > excess cable up and covering with anti-static plastic help? I don't remember there being any specific termination like you see on things like SCSI disks or RL Disks. But then, it has been a long time since I had my last Pertec Tape Drive. Only 9-track I have today is SCSI. bill ' bill
Re: Pertec Interface Cable Length
On 08/11/2019 10:40 AM, Nigel Johnson via cctalk wrote: It's funny how licensing bodies do not recognise computer engineers. I am a member if the IEEE, but since I first wrote to the local body in 1974 they have never recognised computer engineering as a discipline. After twenty years of chip-level troubleshooting on DEC machines I spent twenty twenty-five years teaching college before retiring to my soon-to-be-restored collection of old kit. I ran into the then President of the provincial licensing association at an alumni event a few years ago and he laughed, saying they are still working on it! Well, "computer engineering" isn't well-defined. For EE, you can write loop and node equations and solve, and determine exactly how an electrical network will behave. They are trying to make systems that can analyze computer programs in the same way, but I think we are pretty far from that level of rigor. Jon
Re: Pertec Interface Cable Length
On 8/11/19 9:07 AM, W2HX via cctech wrote: >> Would folding the > excess cable up and covering with anti-static plastic help? > > unlikely. The typical problem is that the longer the cable, the higher the > capacitance of the transmission line. Therefore you get a lot of problems > with rising and trailing edges of the signals which can cause all kinds of > problems if not terminated correctly. I agree with the previous suggestion > to make sure all termination is in place. If you can also place a scope on > one of the lines (with high impedance probe) you can see what the edges look > like. All of my Pertec tape drives (in the past) were located in a differentrack from the actual computer so the cables were always over 10' long and usually close to if not 20'. Never had a problem. I seem to remember they were ribbon cables with each odd/even pair twisted which probably meant one active and one ground twisted together. > > From: cctech on behalf of Douglas Taylor via > cctech > Sent: Sunday, August 11, 2019 12:33 AM > To: Jon Elson; On-Topic Posts > Subject: Re: Pertec Interface Cable Length > > On 8/10/2019 1:56 PM, Jon Elson wrote: >> On 08/09/2019 11:05 PM, Douglas Taylor via cctech wrote: >>> I have a question about cable length - any electrical engineers in >>> the house? >>> >>> Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape >>> controller in a pdp-11/53. The interface is pertec with 2 50 pin >>> cables. >>> >>> When I use a pair of short flat ribbon cables, 18 and 30 inches each, >>> it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. >>> >>> It doesn't work when I use a pair of 5 foot long flat ribbon cables. >>> Are they too long? Do I need twisted pair type of cable? Is it >>> possibly a termination problem? >>> >> I have used cables about 20 feet long without trouble. The 2 50-pin >> cables is the Pertec formatted interface, which is really forgiving. >> Does you drive have terminators in both ends of the cable (both at the >> TC02 end and the drive end)? >> Now, I will mention that I have ONLY used twisted-pair ribbon cables >> with both flavors of interface, never straight ribbon cable. >> >> Jon >> >> > I haven't checked to see if there are terminators (Arnold the > Terminator) on either end. I did check the long cables for continuity > and found no problems. It may be an EMI problem. Would folding the > excess cable up and covering with anti-static plastic help? I don't remember there being any specific termination like you see on things like SCSI disks or RL Disks. But then, it has been a long time since I had my last Pertec Tape Drive. Only 9-track I have today is SCSI. bill ' bill
Re: Pertec Interface Cable Length
Jon Elson via cctalk wrote: On 08/10/2019 01:29 PM, Dave Wade via cctalk wrote: -Original Message- From: cctech On Behalf Of Mark J. Blair What term is used there for an engineer who works in fields of general electronics? An electronics engineer... This war was settled in 1963 when the American Institute of Electrical Engineers merged with the Institute of Radio Engineers, realizing their battle was just silly and counterproductive. It was time, as serious electronics was moving into telecommunications and computers, numerically controlled machine tools, aviation, and more. If they had a separate institute for each area of specialization, it would just dilute the resources. Every one of them used Ohms law and its derivatives. Jon I very much agree with the U.S. philosophy with regards to school organization (i.e., a department or school with general electrical engineering faculty, directly hosting a degree program in electrical engineering, with majors in either communications, power systems, computer engineering, and so on. In the last few decades some departments in the U.S. have transitioned to names such as "Department of Electrical and Computer Engineering" (see e.g. http://www.ece.wisc.edu/ and https://www.ece.cornell.edu/ece, precisely the two universities where I studied in the U.S.). In Latin America, with some exceptions, it is common that each very specific program is hosted by one very specific department, thus there is one electronics engineering program linked to a department of electronics engineering and a different electrical engineering program with a department of electrical engineering. This is especially true in older, public universities. The reason for this is petty: resource management (i.e., handling of tuition, human resources and so on) was often done at the program level, so people preferred to have "their own turf" in the organization and this led to this kind of granularization. I do belong to a "Department of Electrical, Electronics and Computer Engineering" and I like it this way. We do, however, host two differently-named undergraduate programs, called, you guessed it, Electrical Engineering and Electronics Engineering. Nowadays many students are choosing to stay one or two more semesters and obtain the two degrees; this is possible because of the curricular design that we have in place. In the U.S., some of the first Electrical Engineering programs were created inside Physics departments and only later were the corresponding departments created. carlos.
Re: Pertec Interface Cable Length
It's funny how licensing bodies do not recognise computer engineers. I am a member if the IEEE, but since I first wrote to the local body in 1974 they have never recognised computer engineering as a discipline. After twenty years of chip-level troubleshooting on DEC machines I spent twenty twenty-five years teaching college before retiring to my soon-to-be-restored collection of old kit. I ran into the then President of the provincial licensing association at an alumni event a few years ago and he laughed, saying they are still working on it! Meanwhile, computers run everything... cheers, Nigel Johnson On 11/08/2019 11:34, Jon Elson via cctalk wrote: On 08/10/2019 01:29 PM, Dave Wade via cctalk wrote: -Original Message- From: cctech On Behalf Of Mark J. Blair What term is used there for an engineer who works in fields of general electronics? An electronics engineer... This war was settled in 1963 when the American Institute of Electrical Engineers merged with the Institute of Radio Engineers, realizing their battle was just silly and counterproductive. It was time, as serious electronics was moving into telecommunications and computers, numerically controlled machine tools, aviation, and more. If they had a separate institute for each area of specialization, it would just dilute the resources. Every one of them used Ohms law and its derivatives. Jon -- Nigel Johnson MSc., MIEEE VE3ID/G4AJQ/VA3MCU Amateur Radio, the origin of the open-source concept! You can reach me by voice on Skype: TILBURY2591 If time travel ever will be possible, it already is. Ask me again yesterday This e-mail is not and cannot, by its nature, be confidential. En route from me to you, it will pass across the public Internet, easily readable by any number of system administrators along the way. Nigel Johnson Please consider the environment when deciding if you really need to print this message
Re: Pertec Interface Cable Length
On 08/10/2019 01:29 PM, Dave Wade via cctalk wrote: -Original Message- From: cctech On Behalf Of Mark J. Blair What term is used there for an engineer who works in fields of general electronics? An electronics engineer... This war was settled in 1963 when the American Institute of Electrical Engineers merged with the Institute of Radio Engineers, realizing their battle was just silly and counterproductive. It was time, as serious electronics was moving into telecommunications and computers, numerically controlled machine tools, aviation, and more. If they had a separate institute for each area of specialization, it would just dilute the resources. Every one of them used Ohms law and its derivatives. Jon
Re: Pertec Interface Cable Length
>Would folding the excess cable up and covering with anti-static plastic help? unlikely. The typical problem is that the longer the cable, the higher the capacitance of the transmission line. Therefore you get a lot of problems with rising and trailing edges of the signals which can cause all kinds of problems if not terminated correctly. I agree with the previous suggestion to make sure all termination is in place. If you can also place a scope on one of the lines (with high impedance probe) you can see what the edges look like. From: cctech on behalf of Douglas Taylor via cctech Sent: Sunday, August 11, 2019 12:33 AM To: Jon Elson; On-Topic Posts Subject: Re: Pertec Interface Cable Length On 8/10/2019 1:56 PM, Jon Elson wrote: > On 08/09/2019 11:05 PM, Douglas Taylor via cctech wrote: >> I have a question about cable length - any electrical engineers in >> the house? >> >> Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape >> controller in a pdp-11/53. The interface is pertec with 2 50 pin >> cables. >> >> When I use a pair of short flat ribbon cables, 18 and 30 inches each, >> it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. >> >> It doesn't work when I use a pair of 5 foot long flat ribbon cables. >> Are they too long? Do I need twisted pair type of cable? Is it >> possibly a termination problem? >> > I have used cables about 20 feet long without trouble. The 2 50-pin > cables is the Pertec formatted interface, which is really forgiving. > Does you drive have terminators in both ends of the cable (both at the > TC02 end and the drive end)? > Now, I will mention that I have ONLY used twisted-pair ribbon cables > with both flavors of interface, never straight ribbon cable. > > Jon > > I haven't checked to see if there are terminators (Arnold the Terminator) on either end. I did check the long cables for continuity and found no problems. It may be an EMI problem. Would folding the excess cable up and covering with anti-static plastic help? Doug
Re: Pertec Interface Cable Length
On 8/10/2019 1:56 PM, Jon Elson wrote: On 08/09/2019 11:05 PM, Douglas Taylor via cctech wrote: I have a question about cable length - any electrical engineers in the house? Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. When I use a pair of short flat ribbon cables, 18 and 30 inches each, it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they too long? Do I need twisted pair type of cable? Is it possibly a termination problem? I have used cables about 20 feet long without trouble. The 2 50-pin cables is the Pertec formatted interface, which is really forgiving. Does you drive have terminators in both ends of the cable (both at the TC02 end and the drive end)? Now, I will mention that I have ONLY used twisted-pair ribbon cables with both flavors of interface, never straight ribbon cable. Jon I haven't checked to see if there are terminators (Arnold the Terminator) on either end. I did check the long cables for continuity and found no problems. It may be an EMI problem. Would folding the excess cable up and covering with anti-static plastic help? Doug
Re: Pertec Interface Cable Length
On 08/09/2019 11:05 PM, Douglas Taylor via cctech wrote: I have a question about cable length - any electrical engineers in the house? Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. When I use a pair of short flat ribbon cables, 18 and 30 inches each, it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they too long? Do I need twisted pair type of cable? Is it possibly a termination problem? I have used cables about 20 feet long without trouble. The 2 50-pin cables is the Pertec formatted interface, which is really forgiving. Does you drive have terminators in both ends of the cable (both at the TC02 end and the drive end)? Now, I will mention that I have ONLY used twisted-pair ribbon cables with both flavors of interface, never straight ribbon cable. Jon
Re: Pertec Interface Cable Length
> On Aug 10, 2019, at 1:57 AM, Dave Wade via cctech > wrote: > >> >> I have a question about cable length - any electrical engineers in the house? > > Its electronics, rather than electrical engineering. Electrical Engineering > is power distribution. At least in the US, "Electrical Engineering" applies to both subfields. Same university department, same degree name, same generic title of "electrical engineer"; just different specializations. I presume from your remark that an "electrical engineer" in the UK would be an engineer who works in the field of power generation and distribution. What term is used there for an engineer who works in fields of general electronics? -- Mark J. Blair, NF6X http://www.nf6x.net/
RE: Pertec Interface Cable Length
> -Original Message- > From: cctech On Behalf Of Mark J. Blair via > cctech > Sent: 10 August 2019 18:03 > To: General Discussion: On-Topic Posts > Subject: Re: Pertec Interface Cable Length > > > > > On Aug 10, 2019, at 1:57 AM, Dave Wade via cctech > wrote: > > > >> > >> I have a question about cable length - any electrical engineers in the house? > > > > Its electronics, rather than electrical engineering. Electrical Engineering is > power distribution. > > At least in the US, "Electrical Engineering" applies to both subfields. Same > university department, same degree name, same generic title of "electrical > engineer"; just different specializations. I presume from your remark that an > "electrical engineer" in the UK would be an engineer who works in the field of > power generation and distribution. Yes.. > What term is used there for an engineer > who works in fields of general electronics? > An electronics engineer... > > -- > Mark J. Blair, NF6X > http://www.nf6x.net/ Dave G4UGM
Re: Pertec Interface Cable Length
On Sat, Aug 10, 2019 at 9:46 AM Douglas Taylor via cctech wrote: > > I bought the long cables off ebay, so they have to be good? Right? I > think the short cables came from a hamfest. > > The cables can be fairly long, I remember interfacing a TU80 to an > Emulex QT14 (maybe) and the DEC cables were round and about 15 feet > long. And it worked. > > It was too late last night to begin checking the long cables for > continuity, so I fired off the email instead thinking it may be a > termination problem. > > Is it possible for the IDC and Card edge connectors to be put on wrong? > You would want pin 1 to map to pin 1, and so on. > > Doug If you build cables yourself without proper tools it might be easy to end up with bad results. It takes a fair amount of force to properly press a 50-pin connector on to a ribbon cable with even pressure. I picked up a used 3M Scotchflex 3316 manual hand press with locator plates for IDC and card edge connectors a few years back for a reasonable price. It made quick work of properly pressing 50-pin IDC connectors on to ribbon cables when I built some Pertec interface tape drive cables. I used some Amphenol Spectra-Strip Twist 'N' Flat ribbon cable and made a couple of cables that are about 8 feet long. I haven't had any problems using them between an Emulex QT13 controller and a Fujitsu M2444AC tape drive.
Re: Pertec Interface Cable Length
On 8/10/19 9:45 AM, Douglas Taylor via cctech wrote: > I bought the long cables off ebay, so they have to be good? Right? I > think the short cables came from a hamfest. > > The cables can be fairly long, I remember interfacing a TU80 to an > Emulex QT14 (maybe) and the DEC cables were round and about 15 feet > long. And it worked. Of course it did--the TU80 hews to the Pertec inteface spec. The Qualstar, as I observed, does not. It's basically "Pertec on the cheap". It might be interesting if someone with a Qualstar 1xxx series drive who's using 10' flat ribbon cables can report their success. I suspect that you'll hear crickets... The point of the Qualstar drives is that they were cheap--and the design reflects that. --Chuck
Re: Pertec Interface Cable Length
On 8/10/2019 4:57 AM, Dave Wade wrote: -Original Message- From: cctech On Behalf Of Douglas Taylor via cctech Sent: 10 August 2019 05:06 To: General Discussion: On-Topic Posts Subject: Pertec Interface Cable Length I have a question about cable length - any electrical engineers in the house? Its electronics, rather than electrical engineering. Electrical Engineering is power distribution. Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. When I use a pair of short flat ribbon cables, 18 and 30 inches each, it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. So the hardware is good. It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they too long? Do I need twisted pair type of cable? Is it possibly a termination problem? I can't see 5 foot being too long for data from a tape, the data rates aren't huge. At most you have added 10ns to the delay times. On the other hand I have been wrong in the past and could be wrong again.. I assume you have checked the cables. Ribbon cables are prone to come loose from the IDC pins if it’s a IDC connector, and if soldered can break Doug Dave G4UGM I bought the long cables off ebay, so they have to be good? Right? I think the short cables came from a hamfest. The cables can be fairly long, I remember interfacing a TU80 to an Emulex QT14 (maybe) and the DEC cables were round and about 15 feet long. And it worked. It was too late last night to begin checking the long cables for continuity, so I fired off the email instead thinking it may be a termination problem. Is it possible for the IDC and Card edge connectors to be put on wrong? You would want pin 1 to map to pin 1, and so on. Doug
RE: Pertec Interface Cable Length
> -Original Message- > From: cctech On Behalf Of Douglas Taylor > via cctech > Sent: 10 August 2019 05:06 > To: General Discussion: On-Topic Posts > Subject: Pertec Interface Cable Length > > I have a question about cable length - any electrical engineers in the house? Its electronics, rather than electrical engineering. Electrical Engineering is power distribution. > > Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller > in a pdp-11/53. The interface is pertec with 2 50 pin cables. > > When I use a pair of short flat ribbon cables, 18 and 30 inches each, it > works. Under RT11 I can INIT, Copy, DUMP, do a Directory. > So the hardware is good. > It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they > too > long? Do I need twisted pair type of cable? Is it possibly a termination > problem? I can't see 5 foot being too long for data from a tape, the data rates aren't huge. At most you have added 10ns to the delay times. On the other hand I have been wrong in the past and could be wrong again.. I assume you have checked the cables. Ribbon cables are prone to come loose from the IDC pins if it’s a IDC connector, and if soldered can break > > Doug > Dave G4UGM
Pertec Interface Cable Length
I have a question about cable length - any electrical engineers in the house? Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. When I use a pair of short flat ribbon cables, 18 and 30 inches each, it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they too long? Do I need twisted pair type of cable? Is it possibly a termination problem? Doug
Re: Pertec Interface Cable Length
> It doesn't work when I use a pair of 5 foot long flat ribbon cables. > Are they too long? Do I need twisted pair type of cable? Is it > possibly a termination problem? These cables typically were twisted pair, at least inside cabinets. (They might transition to round cables at bulkheads before routing to other cabinets.) I've seen them in lengths of like 20-30 feet. De
Re: Pertec Interface Cable Length
On 8/9/19 9:05 PM, Douglas Taylor via cctech wrote: > I have a question about cable length - any electrical engineers in the > house? > > Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape > controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. > > When I use a pair of short flat ribbon cables, 18 and 30 inches each, it > works. Under RT11 I can INIT, Copy, DUMP, do a Directory. > > It doesn't work when I use a pair of 5 foot long flat ribbon cables. > Are they too long? Do I need twisted pair type of cable? Is it > possibly a termination problem? 5 feet should be no problem--ISTR that Pertec specifies a maximum cable length of 20 ft. I use 10 foot cables routinely. The Pertec interface is basically open-collector drivers into 220/330 ohms at the far end. I have a Qalstar 1260, but it's the 1260S, the SCSI version--and I rarely use it. Looking at the schematics of the similar Qualstar 1052, the output drivers are plain old LS240s; 24 ma totem-pole outputs. The usual practice is to use 7438s OC 48 ma outputs; it's certainly the case for Pertec formatters. So it could be that you're limited by the Qualstar design. Twisted flat cable might buy you some added distance or you could fabricate a "repeater" with real OC drivers to put between the drive and the controller to extend the length. FWIW, Chuck
Re: Pertec Interface Cable Length
On 8/9/19 10:05 PM, Douglas Taylor via cctech wrote: I have a question about cable length - any electrical engineers in the house? I'm not one. Connected a Qualstar 1260 tape drive to an Emulex TC02 qbus tape controller in a pdp-11/53. The interface is pertec with 2 50 pin cables. When I use a pair of short flat ribbon cables, 18 and 30 inches each, it works. Under RT11 I can INIT, Copy, DUMP, do a Directory. It doesn't work when I use a pair of 5 foot long flat ribbon cables. Are they too long? Do I need twisted pair type of cable? Is it possibly a termination problem? I've heard tell that flat ribbon cables that long don't work for much at any speed worth while. I've also heard tell that you can twist the cable and likely help it to the point that it might work. I'd try for a twist every 1–2 feed. I'd also likely try twisting the cables in the opposite direction. This opinion comes with a 100% money back guarantee. 100% is opinion, guaranteed! It's free to try, and it might work. So maybe give it a try. Here's hoping someone else with more information will respond too. -- Grant. . . . unix || die
