Print

[W1QJ]

Looking at an Amp Supply LK800 with QSK board, probably same for the other models too with QSK.  The RJ1 vacuum relay coil is a bit flaky and I’d like to install a new one.  It’s stamped 12 on it.  I assume that’s the 12v coil.  I don’t see in the amp where more than 12volts is developed.  I removed the coil and measured about 340 ohms which is what the 26.5 coil measures.  The 12v coils usually measure about 85 ohms.  The manual does not say what voltage the vacuum relay is.  All the other relays on that board are clearly marked 12volts and the schematic shows them in parallel.   Because of the conflict I’d like to know if anyone knows with certainty what the coil voltage is supposed to be.  Thanks for any help.

[VE7RF]

Looking at an Amp Supply LK800 with QSK board, probably same for the other models too with QSK.  The RJ1 vacuum relay coil is a bit flaky and I’d like to install a new one.  It’s stamped 12 on it.  I assume that’s the 12v coil.  I don’t see in the amp where more than 12volts is developed.  I removed the coil and measured about 340 ohms which is what the 26.5 coil measures.  The 12v coils usually measure about 85 ohms.  The manual does not say what voltage the vacuum relay is.  All the other relays on that board are clearly marked 12volts and the schematic shows them in parallel.   Because of the conflict I’d like to know if anyone knows with certainty what the coil voltage is supposed to be.  Thanks for any help.

 26.5 vdc  coils are sup[posed to be aprx 335 ohms.  The 12 vdc coils are supposed to be 80 ohms.  The coils are field replaceable in either  12/ 26.5 vdc versions.  It might be u have a 26.5 vdc coil in there...that happens to operate ( barely) with 12 vdc applied. I have a bunch of jennings rj1a's  with 26.5 vdc  coils.  When I have some time, I will use the lab supply, and bring the vdc up slowly..and see what the exact pick up vdc is.  If it does operate with 12 vdc applied, it's operate speed is gonna be slow, like 8 msecs. With 26.5 vdc applied to a 26.5 vdc coil, speed is typ 2-3 msecs.

BTW, where is KM1H these last several months    He has just up and vanished into thin air.   Is he ok ?

[K0IZ]

My LK800C has the vacuum relay fed from 20VAC, through 1/2wave rectifier (diode) and filter cap.  So x 1.4 is about 28 volts or so.  Do you need schematic?

[W1QJ]

K0IZ I would like to see your schematic thank you.  Lou

Jim, I have no idea about Carl and don’t know anyone that does.

[KD6VXI]

Carl is OK.  A few people on Eham had the local PD do a health and welfare check.  Carl basically said "I'm alive".  I have no reason to think he's not OK at this time.  I think, like JI, he pretty much got tired of having to argue information into peoples heads and decided living was more important.  A shame.  He's gruff, but he's forgotten more than most people on this reflector will know about radio.

I have used 26 volt vac relays of various manufacturers with 14.x volts before.  Possibly 15 volts floating, pulled down.  The 26 volt relays need a slightly higher voltage to kick, but once they do, a lower voltage will hold them in.  As such, small amounts of C and a half wave will usually work.

Lou, you can't measure the voltage in the amp?

--Shane
WP2ASS / ex KD6VXI

[W1QJ]

Shane, yes I could, but that doesn’t take away some remote chance that a higher voltage coil was used for some reason.  K0IZ sent me a schematic of a QSK LK800 he has and it is clearly marked with a 20vac input AC into a half wave rectifier using a 26.5v relay. It also has a 555 chip on the board.  My unit along with the schematic shows no 555 chip and no 20vac input.  Mine is different for sure and it’s all 12v.  Apparently different renditions of that QSK board.  There’s no doubt at this point.  Thanks to John for sending the schematic to my email address. 

[VE7RF]

Shane, yes I could, but that doesn’t take away some remote chance that a higher voltage coil was used for some reason.  K0IZ sent me a schematic of a QSK LK800 he has and it is clearly marked with a 20vac input AC into a half wave rectifier using a 26.5v relay. It also has a 555 chip on the board.  My unit along with the schematic shows no 555 chip and no 20vac input.  Mine is different for sure and it’s all 12v.  Apparently different renditions of that QSK board.  There’s no doubt at this point.  Thanks to John for sending the schematic to my email address.

RL Drake used a 22 vac winding (on the 5V@ 30 amp fil xfmr)..... then a 1/2 wave rectifier.  Dentron also used a 1/2 wave rectifier, WTF ?  The 1st thing I do is toss any 1/2 wave rectifier and replace with a FWB.... but I guess they tried to...'save' money. Any  1/2 wave rectifier is a dumb idea at the best of times.  It only conducts on 1/2 cycles. On the other 1/2 cycle, full rectified DC is flowing through the xfmr secondary winding.  Converting to a FWB means you can now pull double the DC current from the same supply.

But this probably doesn't help Lou... unless they obtained the 12 vdc from a 1/2 wave rectifier.  The RJ1A's coil can be field swapped with either the 12 vdc coil (80 ohms) or the 26.5 vdc coil (335 ohms).   When you do that, the cover plate is reversed to match the actual coil voltage. (It's marked 12 vdc / 80 ohms on one side...... and 26.5 vdc / 335 ohms on the other side).  Along the line some where, I suspect the oem 12 vdc relay was swapped to what they thought was another 12 vdc relay... which actually has a 26.5 vdc coil.

The 26.5 vdc coil will operate with 12-13 vdc ( pick up voltage).....and will RLS  (drop out voltage)... with 1-10 vdc.
When it just barely operates like that, it's typ 8 msecs.  When run at rated voltage, it's typ just 2-3 msecs.  With any amount of over voltage, it's  1 msec, or slightly less.

Jennings no longer exists, but you can still buy the 12 / 26.5 vdc coils from max gain systems.  The gigavac GH-1 is a direct swap, but beware, the threads on the GH-1 are metric, as is the mating lock nuts.  The GH-1 also comes in 12 vdc / 26.5 vdc versions, and again, coils are field replaceable. The coils are seldom a problem. But for some applications, one voltage may be preferred over the other.  Kilovac engineers told me that if you want to speed em up, and are voltage limited, to use the 12 vdc coil.... ( + correct drop resistor).

[AE0Q]

Kilovac engineers told me that if you want to speed em up, and are voltage limited, to use the 12 vdc coil.... ( + correct drop resistor).

The Alpha 77D/Dx/Sx amps, also 91B and some 99 used RJ1A relays with a 26.5 vdc coil.  The later 99, 8100 and 8410 amps are Kilovac relays, same voltage.  Those amps all use a speed-up (ETO called it a 'hot-shot') circuit to make them fast on the initial closing.  It initially applies the 40vdc Aux voltage to the coil via a dropping resistor and capacitor, then once the relay is keyed the voltage across the coil is the correct 26 vdc.  The circuit is explained in the 77Dx/Sx manual.

[W1QJ]

The saga seems to have been resolved.  First off, this amp had the golden screwdriver treatment before I got it.  There seems to be an old version and new version QSK board for these amps.  Apparently it seems the amp came with an old version QSK board that  worked with a 12v system.  The old filament transformer had a 12v or  less winding for 12 relays etc.  someone apparently installed a later model QSK board that used a 20vac input.  Since the older filament transformer didn’t have the 20v winding someone installed a Radio Shack 12v transformer which gives about 17vdc.  They apparently opted for a 12v instead of the 26.5v coil.  The coil was intermittent so I replaced it with another 12v coil and it’s working fine.  I can always add a resistor if I have to.  I’m just happy it’s keying up. 

[VE7RF]

The saga seems to have been resolved.  First off, this amp had the golden screwdriver treatment before I got it.  There seems to be an old version and new version QSK board for these amps.  Apparently it seems the amp came with an old version QSK board that  worked with a 12v system.  The old filament transformer had a 12v or  less winding for 12 relays etc.  someone apparently installed a later model QSK board that used a 20vac input.  Since the older filament transformer didn’t have the 20v winding someone installed a Radio Shack 12v transformer which gives about 17vdc.  They apparently opted for a 12v instead of the 26.5v coil.  The coil was intermittent so I replaced it with another 12v coil and it’s working fine.  I can always add a resistor if I have to.  I’m just happy it’s keying up.

I would install the 33.33 ohm drop resistor in series with the 17 vdc.  It will only dissipate .75 watt. To speed it up some more, install a 50 uf cap across the 33.33 ohm resistor.   With 50 uf, the RC time constant is just 1.66 msecs.   You need  2-3 time constants to discharge the cap.  ( 50 uf = .00005 farad.  .00005 X 33.33 = .001666 secs, or 1.666 msecs.

Without the 33.33 ohm resistor installed, the 12 vdc coil will draw 213 ma.... instead of 150 ma.  With 215 ma, you may well eventually trash the new coil.    30 ohms will suffice, and can be made from 2 x 15 ohms in series, or 2 x 60 ohms in parallel.

Jennings + Kilovac  called the resistor + paralleled cap combo a .... 'RC hotshot' speed up circuit.  That concept has been around since the 50's..and works well. Just don't get the time constant too high for QSK  CW use.  It can be a lot longer for VOX SSB / CW use...like 10 msecs.  When T=0, the discharged cap looks like a dead short, which also shunts the series drop resistor it's in parallel with.   The full  OCV (open circuit voltage...in this case, 17 vdc)  is applied to the coil. Once the cap is fully charged up, 150 ma will then flow through the 33.33 ohms series drop resistor... and the normal 12 vdc is applied to the coil...and the remaining 5 vdc is dropped across the drop resistor.  Caps block DCV of course, so once fully charged up, the only other path is then through the drop resistor.

[W1QJ]

Thanks Jim, I’ll do an exact measurement of the voltage and see just what it is exactly.  I just calculated it but haven’t measured it.  I’ll do that.

[VE7RF]

Kilovac engineers told me that if you want to speed em up, and are voltage limited, to use the 12 vdc coil.... ( + correct drop resistor).

The Alpha 77D/Dx/Sx amps, also 91B and some 99 used RJ1A relays with a 26.5 vdc coil.  The later 99, 8100 and 8410 amps are Kilovac relays, same voltage.  Those amps all use a speed-up (ETO called it a 'hot-shot') circuit to make them fast on the initial closing.  It initially applies the 40vdc Aux voltage to the coil via a dropping resistor and capacitor, then once the relay is keyed the voltage across the coil is the correct 26 vdc.  The circuit is explained in the 77Dx/Sx manual.

AFAIK, the 9500 also uses a kilovac HC-1 relay.  How come the switch from pin diodes, like used in the 87A + 89 ??
Alpha started off with a jennings RJ1A....then later switched to the more robust Kilovac HC-1..then switched to PIN diodes...than back to a HC-1.  On a side note, the UK  SS amp also uses a Kilovac HC-1.  They rate it at 50 wpm... which I find questionable.  I'm not a QSK  CW op, but these vac relays must take a beating if used in QSK CW mode, like in a 48 hr contest. K9YC has replaced the vac relay in his Ten-Tec amp, several times, in the middle of a CW contest...used with his K3, and everything sequenced correctly. He gave up on replacing vac relays, and now uses a KPA-1500..which uses pin diodes.   

[AE0Q]

AFAIK, the 9500 also uses a kilovac HC-1 relay.

No, the 9500 uses the Gigavac (G41C334) or Kunshan (V41C334) vacuum relay (as shown on the Alpha web site).  They have a narrower body and higher voltage/current rating.

There is a web page (not Alpha's) showing a modification to mount RJ1A-type relays in the 9500, inspired to lower the relay noise.

How come the switch from pin diodes, like used in the 87A + 89 ??

As I answered on Oct 23:

My guess is the high power PIN diodes cost a lot AND had/have to be special ordered in large batches from one company.  The diodes for the 87A/89 are now unavailable, although you can stuff slightly bigger ones in there.
You would have to ask Gordon W0RUN his exact reasons for the 9500 design..

[AC2RY]

Alpha started off with a jennings RJ1A....then later switched to the more robust Kilovac HC-1..then switched to PIN diodes...than back to a HC-1.  On a side note, the UK  SS amp also uses a Kilovac HC-1.  They rate it at 50 wpm... which I find questionable.  I'm not a QSK  CW op, but these vac relays must take a beating if used in QSK CW mode, like in a 48 hr contest. K9YC has replaced the vac relay in his Ten-Tec amp, several times, in the middle of a CW contest...used with his K3, and everything sequenced correctly. He gave up on replacing vac relays, and now uses a KPA-1500..which uses pin diodes.

Relays are rated for 2M cycles. This is not very much if full QSK is used. Likely need to replace after few years of service if use that mode daily.

[AE0Q]

Relays are rated for 2M cycles. This is not very much if full QSK is used. Likely need to replace after few years of service if use that mode daily.

Years ago I added up the cycles needed to make a QSO in the CQWW CW contest, if you were a run station.  A vacuum relay ended up being rated for about 30,000 contest contacts.