I did a little testing today with a KXN1052 UHF element at 448.125MHz. As previously mentioned, a crystal manufactured by Bomar cut for 448.125-11.7/12 was installed, and the resulting carrier frequency was almost 15KHz high. Adjustment of the trimmer pulled the crystal within 4KHz high of the operating frequency. I then padded a 10pf capacitor in parallel with the trimmer cap C6. This lowered the operating frequency such that the trimmer would allow a swing of 10KHz below and 20KHz above the desired frequency. However, as previously reported, the element would only allow a maximum deviation of +/- 3.5KHz with the IDC adjustment at maximum setting.
I then remembered that years ago I went thru this exercise before, and had some luck experimenting with random changes. At the time there was no schematic of the element available. This time, armed with the KXN1019 schematic on RBTIP (closest since there was no KXN1052 schematic posted yet), I wanted to try to figure out how the major crystal mfgs were "compensating" their elements to work with their crystals. I first removed C1, a 36pf +/-5% ceramic disc capacitor, which resulted in the frequency dropping way too low, almost 20KHz below the desired carrier frequency. Since the 36pf cap resulted in an element 20KHz to high, and no C1 in place resulted in an element 20KHz too low, I decided to experiment by tack-soldering different value caps in place and trying the element. I found an old 15pf cap that was identical in style (came from a Micor VHF rx rebanded to 2m) and temp coefficient value. The new cap resulted in the C6 trimmer being able to swing the crystal between -10KHz and +15KHz of the desired carrier frequency. And the best part - the element was able to deviate upwards of 8KHz with the IDC at max setting. I permanently soldered the new cap properly in place, and re-assembled the element. So, my conlusion is that the crystal mfgs must be playing with the value of C1 to "compensate" for their crystal cuts. It appears that there is enough room in the value of C1 to "compensate" for various cut crystals. Looking at the schematic, it appears C2 and C3 (both 20pf) may also be modified with similar results, but without the element in front of me, I can't recall if they were ceramic disc caps or some other type. I don't have an element handy at the moment. However, they are in parallel with C1. Looking further at the schematic, I am guessing that the temperature compensation is being controlled by C4 & C5, which are each in series with a thermistor, then paralleld with C1, C2, & C3. Since they weren't touched, and my replacement C1 was of the same series, I would deduct that I should have similar temp compensation, sans any difference between the Bomar crystal and the original Motorola crystal. Note: The part locations/designations above are from the KXN1019 schematic; the placement of the parts mentioned above is almost identical to the KXN1052 I was working with. Comments? Eric KE2D --- In [email protected], "Gary Schafer" <[EMAIL PROTECTED]> wrote: > > I don't know about all crystal companies but some like it when you send them > a crystal that works properly in your circuit and they then duplicate the > new crystal to work in that same circuit. > > So if you sent them just the crystal from a channel element that was working > properly they would make you a new crystal to work in that circuit on the > frequency of your choice. > > You may be able to get by with just sending the crystal. I would question > how much temperature compensation they actually do to the channel element as > that is a long drawn out process. My guess is they just build the crystal to > work with the existing parameters of the channel element. > > 73 > Gary K4FMX > > > -----Original Message----- > > From: [email protected] [mailto:Repeater- > > [EMAIL PROTECTED] On Behalf Of Eric Lemmon > > Sent: Friday, March 30, 2007 12:16 AM > > To: [email protected] > > Subject: RE: [Repeater-Builder] Re: ICM Crystals Off Frequency? > > > > To the best of my knowledge, the crystal is always manufactured first to > > meet ICM's nominal specifications, and the channel element is then > > modified > > as necessary to perform satisfactorily with that crystal. > > > > 73, Eric Lemmon WB6FLY > > > > > > -----Original Message----- > > From: [email protected] > > [mailto:[EMAIL PROTECTED] On Behalf Of kk2ed > > Sent: Thursday, March 29, 2007 9:04 PM > > To: [email protected] > > Subject: [Repeater-Builder] Re: ICM Crystals Off Frequency? > > > > Ok, but back to my original question - when ICM does an element - do > > they cut the crystal first to "their standards", then modify the > > element to make it work? Or are they measuring the element first in > > some fashion, then cutting the crystal to the "known" element > > characteristics? > > > > --- In [email protected] > > <mailto:Repeater-Builder%40yahoogroups.com> , "Eric Lemmon" <wb6fly@> > > wrote: > > > > > > Each channel element, regardless of the manufacturer, contains a > > number of > > > resistors, capacitors, and perhaps a few inductors. The capacitors > > have > > > specific TCs (Temperature Coefficients) that are chosen so that the > > > capacitance variation with temperature change is exactly > > complementary to > > > the reaction of the crystal. When performed correctly, the > > capacitors > > > change value with temperature just enough to cancel out the > > frequency drift > > > of the crystal. > > > > > > However, a full compensation of the crystal holder (channel > > element, ICOM, > > > etc.) includes more than temperature compensation. The technician > > also > > > verifies that the crystal can be set exactly on frequency with the > > included > > > trimmer, that the output amplitude meets the minimum specification, > > and that > > > the crystal is "rubbery" enough to be modulated to the required > > deviation > > > level. > > > > > > As you might expect, full compensation of a channel element to a > > particular > > > crystal is an exacting and time-consuming process. That's why ICM > > charges > > > more for the compensation than the crystal costs. > > > > > > When a radio user orders just the crystal and puts it into a handy > > channel > > > element, the components inside that channel element may or may not > > match the > > > characteristics of the new crystal. As you and many others have > > discovered, > > > the new crystal may be such a poor match to the channel element that > > it may be > > > impossible to get it to operate on frequency. Even if you can add > > or remove > > > some shunt capacitance to tweak the crystal onto frequency, that > > shunt > > > capacitance is not temperature compensated. It may work fine, and > > it may > > > not. > > > > > > Both Motorola and General Electric operated their own crystal > > manufacturing > > > facilities for many years. Since each company had complete control > > over the > > > making of both the crystal and the channel element that contained > > it, they > > > could evolve the processes to optimize performance and longevity. > > Let's say > > > that Motorola found that their MICOR channel elements worked best > > with > > > crystals that were made for a 25 pF load rather than a 30 pF load. > > If you > > > have one of these original MICOR channel elements that you want to > > > re-crystal, it is likely that ICM or Bomar or Crystek will ship you > > a > > > nominal crystal, since they have no way of knowing that your > > channel element > > > is not nominal but has already been compensated to the original > > crystal, > > > which may have a non-nominal load capacitance. How can they know, > > if you > > > don't send in the channel element? Also, since the crystal house > > never had > > > the chance to test your channel element first, they have no > > obligation to > > > make changes to your crystal if it doesn't work properly once you > > install > > > it. > > > > > > Given that a full compensation is a one-time charge, I personally > > have every > > > crystal I buy given the full compensation in a channel element I > > send to the > > > crystal house. I think it's a prudent investment. Not everyone > > agrees... > > > > > > 73, Eric Lemmon WB6FLY > > > > > > > > > -----Original Message----- > > > From: [email protected] > > <mailto:Repeater-Builder%40yahoogroups.com> > > > [mailto:[email protected] > > <mailto:Repeater-Builder%40yahoogroups.com> ] On Behalf Of kk2ed > > > Sent: Thursday, March 29, 2007 7:49 PM > > > To: [email protected] > > <mailto:Repeater-Builder%40yahoogroups.com> > > > Subject: [Norton AntiSpam] [Repeater-Builder] Re: ICM Crystals Off > > > Frequency? > > > > > > > > > Here's a question I haven't seen answered, but I'm sure a few of us > > > would like to see answered - > > > > > > What exactly is ICM and others doing when they "compensate" or > > match > > > an element with a crystal to get it netted on freq? > > > > > > I've had mixed success. Some crystals & elements tune on freq just > > > fine, while some don't. I've had mixed luck padding extra > > > capacitance on Micor elements, but Mitrek elements use the inductor > > > instead. > > > > > > For example, I have a KXN1052 with a crystal that is 20KHz high > > after > > > dropping it in. I can pad the trimmer, but then the element won't > > > produce more than 3KHz deviation max. > > > > > > > > > What's their secret? The only thing in the element is resistors and > > > capacitors! I'm sure those of us capable of working on a repeater > > are > > > capable of changing a few components. > > > > > > Anyone??? > > > > > > > > > > > > > > > > > > > > > > > > > > > Yahoo! Groups Links > > > > > > >
