Print

[VE7RF]

Meanwhile, thousands of  SB-220's are  still chugging away.

None of which offer the flexibility of, say, an IC-2KL/AT-500 pair when run with a corresponding Icom transceiver. Or a K3 and KPA1500.

I've put many miles on tube amps of all permutations (including some homebuilt contraptions) and none offered the ability to rapidly change bands then automatically tune the antenna system for optimum power transfer. Not a big deal if you only operate a couple of frequencies on two bands, using one amp per...but a very big deal if you're all over the HF spectrum with only one station.

My tube amps will keep on amplifying - I don't intend to sell them - but at the same time they aren't the first choice for the majority of my DX hunting endeavors. Much as my 6146-based hybrid transceivers are viewed: Some are great performers but the solid-state gear in use here snags the majority of DX or contest points.

I don't user tuners any more.  Vac relays on the 40m yagi, and a seco systems  compressible coil setup, with pre-sets for the 80m  rotary dipole.   The  20-10m  (unused) yagis  don't require a tuner.

The K3 has lousy  TX  IMD...at just  -29db  pep (-21 dbc).  That got scratched off my list a long time ago.   After  repairing and aligning, and operating  several  Icom  xcvrs',  I won't have one in the shack,  they don't do what I want.

Dunno why the rush to change bands.  We tested how long it took to  re-tune buddy's  3x6  hb amp, which uses a bandswitch +  2 x vac caps +  mating turns counters.....  + a manually tuned, tuned input, with 2 x vernier driven air caps, and yet another bandswitch.   From top end of  15m..to the bottom end of  160m, was exactly  45 secs.   That's using the cheat sheet, and dialing up pre-sets.   That amp only covers  160-80-75-60-40-30-20-17-15m (no 12 + 10).   45 secs is a worse case.  Typ is upper bands  during the day for most folks, and lower bands at night.

I have several new jenning  motor driven caps ( 10-1000 pf @  15 kv  ceramic) that can be used with a stepper motor for the bandswitch on the next project,  but have not decided..yet.

[HAMHOCK75]

The grain of wheat light bulb behind the meter burned out. It was replaced with a white LED soldered with clear heatshrink tubing shown below and a 1.8 k resistor to drop the 13.8 VDC with about 6.7 mA current.



The color looks like this which compared to photos of the 2KL is changed from a warm light orange to slightly green.



This is a photo of the anti-vibration mounts more commonly known as Wells or expansion nuts. In the photo, the last remaining of the old mounts is shown on the left.



What made it confusing while looking for these mounts is that they look quite different when new as shown here with the new part on the left.

[W9IQ]

If you would like a more even lighting of the meter face, take a look at the single LED light pipes used for LCD backlighting. I have found that they conveniently fit nicely along the top edge of the meter housing and provide a nicely diffused light.

I am surprised they consider the blind buts to be anti-vibration. Do they really fulfill that purpose? It would seem more likely that they are used to avoid press fit nuts (or tapping if sufficiently thick) in the interior sheetmetal pieces.

- Glenn W9IQ

[HAMHOCK75]

The LED used above is a standard viewing angle device, probably about 60 degrees. Normally I use a 140 degree LED described below but discovered the cupboard bare last night.



Some manufacturers used a piece of paper in front to diffuse the light. This is what a a 1/2″ strip of paper does to the display. You can see the actual color of the LED above the meter.



I will order some more wide angle LED's and a light pipe to try out.

I was not sure why the 2KL needs vibration damping. My guess is that the original fan from 40 years ago might have created more noise and vibration than current fans. I noticed that the 40 cfm computer fan draws more current than the newer 50 cfm that replaced it. I didn’t notice any excessive vibration from either fan before the Wells nuts came in. It did seem to make sense if fan vibration was Icom’s concern, to rigidly mount the fan to the heavy heat sink, heat pipe, amplifier assembly to increase the mass, then damp the entire assembly.

[KD8MJR]

Dunno why the rush to change bands.  We tested how long it took to  re-tune buddy's  3x6  hb amp, which uses a bandswitch +  2 x vac caps +  mating turns counters.....  + a manually tuned, tuned input, with 2 x vernier driven air caps, and yet another bandswitch.   From top end of  15m..to the bottom end of  160m, was exactly  45 secs.   That's using the cheat sheet, and dialing up pre-sets.   That amp only covers  160-80-75-60-40-30-20-17-15m (no 12 + 10).   45 secs is a worse case.  Typ is upper bands  during the day for most folks, and lower bands at night.

It's kind of funny you say that because I have lost count of the amount of times I have caught a needed entity on a band just at the point when they are either just calling CQ or when they are saying one more, one more. 

Almost every time I was able to fire up the THP in about 5 seconds and get through.  I do remember one time a station in India called CQ DX and I hesitated so I could hear the call sign a second time and of course someone else got through first.  I thought no problem I will just wait until his contact is complete.  Holy cow, the moment the guy said 73s it was like a wall of calls before I could even key my mike.  When that cluster gets a rare spot posted and then the first guy talks for 1-2 minutes its no longer shooting fish in a barrel.

[VE7RF]

Dunno why the rush to change bands.  We tested how long it took to  re-tune buddy's  3x6  hb amp, which uses a bandswitch +  2 x vac caps +  mating turns counters.....  + a manually tuned, tuned input, with 2 x vernier driven air caps, and yet another bandswitch.   From top end of  15m..to the bottom end of  160m, was exactly  45 secs.   That's using the cheat sheet, and dialing up pre-sets.   That amp only covers  160-80-75-60-40-30-20-17-15m (no 12 + 10).   45 secs is a worse case.  Typ is upper bands  during the day for most folks, and lower bands at night.

It's kind of funny you say that because I have lost count of the amount of times I have caught a needed entity on a band just at the point when they are either just calling CQ or when they are saying one more, one more. 

Almost every time I was able to fire up the THP in about 5 seconds and get through.  I do remember one time a station in India called CQ DX and I hesitated so I could hear the call sign a second time and of course someone else got through first.  I thought no problem I will just wait until his contact is complete.  Holy cow, the moment the guy said 73s it was like a wall of calls before I could even key my mike.  When that cluster gets a rare spot posted and then the first guy talks for 1-2 minutes its no longer shooting fish in a barrel.

Point taken.   Tell that to the folks with the dreaded 3 min fil warmup.......or  8 mins for the  YC-156 / YC-172. The newer  SS  amps  take 10-15-20  secs  to  fully boot up.

[KM1H]

It's kind of funny you say that because I have lost count of the amount of times I have caught a needed entity on a band just at the point when they are either just calling CQ or when they are saying one more, one more.

Its kind of funny that Ive never had that problem with the LK-500 Ive had since new in 1986 and still runs the original Eimac 3-500Z's. A chart with settings for all bands and CW/SSB can be used with one hand and the power switch with the other.
No rocket science needed.

Carl

[VE7RF]

It's kind of funny you say that because I have lost count of the amount of times I have caught a needed entity on a band just at the point when they are either just calling CQ or when they are saying one more, one more.

Its kind of funny that Ive never had that problem with the LK-500 Ive had since new in 1986 and still runs the original Eimac 3-500Z's. A chart with settings for all bands and CW/SSB can be used with one hand and the power switch with the other.
No rocket science needed.

Carl

  IE:  .....'dial it up by the numbers'

[HAMHOCK75]

The accessory 24 pin connector was found to have a broken mounting ear. The connector was held in place with clear epoxy all around the perimeter but did not hold. The epoxy came off in large flakes. To repair this required a solution that would allow the ear to flex if the connector were to be remain removable in the normal fashion.



I decided to try a stainless steel mesh from a plastic welding kit and a sacrificial Molex connector to supply additional plastic if needed to cover the mesh. The mesh is embedded in the existing plastic with a hot soldering iron tip including the joint so the stainless steel provides all the strength.



To make it more difficult to pull the mesh out, the mesh was wrapped around the rear of the ear.  Covering the mesh with more plastic did not turn out well. The additional plastic tended to burn black easily.



Not very pretty but seems strong enough. Here it has been reinstalled in the recessed mounting bracket.



I replaced the HP6291A with an HP6274B lab supply which can supply +40 VDC at 18 A. With this change, a quick test across all bands with the Icom IC-7100 set for about 12 watts output gave this result. The gain variation is about 2 dB. No ALC was used.



This is the power output versus power input as the output power of the Icom IC-7100 is varied from about 1.38 to 16.30 watts out.



A +40 VDC, 30 A. switching power supply is on order. It may arrive next week.

[HAMHOCK75]

The 40 VDC, 30 A switching power supply was delivered. It was tested no load at +40.16 VDC. It was hard wired to the 2KL since the rear panel power connector has not come in yet. With the switching power supply I was able to get 500 watts output on all bands at about 22.5 A.

One concern about the power supply fell by the wayside. The Icom IC-2KLPS has two relays to sequentially turn on so the inrush current is limited. With this switcher plugged into an outlet strip with the Icom IC-7100, it’s power supply, and a power meter, turn on did not result in the circuit breaker tripping. This 1.2 kW power supply was $76 from Aliexpress. Total was $102 with freight and tax.

It did suffer some damage in shipment. Notice the bend in the upper left.



The voltage from the power supply was measured at not transmitting load ( +40.09 VDC ) to full load at 22.5 A ( +40.02 VDC ). Below the efficiency of the 2KL is plotted up to about 500 watts output against the efficiency of theoretical class A and B amplifiers. Increasing bias current pushes the operation down from the green class B line towards the blue class A line.

[HAMHOCK75]

It is occasionally suggested that only switching power supplies made for amateur use are sufficiently free of spurious products. The following  tests where done on this power supply to see if it has adequate spurious suppression. The output is about 900 watts DC to the 2KL for 500 watts RF output at 3.6 MHz in the following measurements.

The first test was to find the switching frequency in the time domain. A x100 scope probe with the ground connection shorted to the input was used as a pickup for a digital scope which resulted in the display below showing a frequency at about 83 kHz. The scope capture does not show that the frequency fluctuates quite bit.



To show the dispersion in the frequency, a spectrum analyzer was used with a x1 scope probe with the ground connection terminating the input again as a probe. The result below shows what looks like an FM spectrum at about 85 kHz with harmonic spectrums at twice and three times the fundamental.



Next the 500 watt output of the 2KL was measured at 3.6 MHz via about a 32 dB coupler plus an additional 10 dB, 2 watt pad in front of the spectrum analyzer.

The first scan was in a 500 kHz span which should show a the 1st, 2nd, and 3rd harmonics of the switching frequency but nothing was visible at over 70 dB below the carrier.



The span was then narrowed to 200 kHz in which any fundamental of the switching frequency at about +/- 85 kHz should be visible but again nothing was found at over 70 dB below the carrier.

[ZS5WC]

 ;)Great thread!..
I have an umolested original IC2KL with its matching supply.
Had it for almost 15 years now.
Rigs come and go but 2KL stays. As mentioned before ICOM should never have discontinued this amp, maybe just modernised it.

My ONLY Gripe was the Electrolux fan noise when it got hot!.. Like a Jet drag car at full tilt!.
I carefully took it apart-the bottom cover acts as the sub chassis so that was a bit of a surprise, and replaced the noisy fan with a suitable modern plastic unit.
I also then applied my own version of the fan mod-On TX the fan supply is switched from the TX line through the Clixon temp. switch to fan +.
So I interrupted this supply, added 1n4007 diode from TX line to Clixon and over clixon I about 47ohm Resistor, 2w.
From the +12v meter supply line added another 1N4007 diode and about 100Ohm 2w in series to feed power to fan when 2KL is on.
Diodes prevent back feed to meter lamp supply and +TX lines.
So Now fan comes on slowly when 2KL is on, runs slightly faster on TX, and if need be will run full speed if heatsink is hot.
I don't HEAR the fan, there is almost no fan noise on TX--I have to listen carefully.., and FULL speed has never been needed.

I left the meter lamp as is, but in my IC765 that uses same rubber mount and reflector for the lamp, I used an 5mm clear super bright led.
I used a scotch pad and dulled the lens a bit..
Works 100% and is WAY brighter than the original lamp.

73 de William
ZS4L / ZS5WC

[HAMHOCK75]

The new wide angle LED’s came in yesterday.



This is what the meter looks like now with the wide angle LED and about a 5/8” wide strip of paper vertically in front of the LED.



Parts that have not yet come in are the rear panel mating power connector and the LCD light pipe. I also need to see if the switching power supply has enough room in it for a relay so the 2KL front panel Power ON/OFF switch can control it like with the 2KLPS.

[KM1H]

This is what the meter looks like now with the wide angle LED and about a 5/8” wide strip of paper vertically in front of the LED.

All that from one LED....WOW. I can see them in some of my boatanchor receivers where replacing the 47's, etc can be a PITA and reducing voltage doesnt look good.

[HAMHOCK75]

Carl

Yes, just the one LED. This photo shows the back side of the meter with the strip of paper providing attenuation directly in front of the LED due to the LED holder being so close to the center of the meter. The paper is held in place with a strip of clear cellophane tape over the copper color metal that the paper is resting on.

The only difference between the meter photo in reply #33 and reply #42 is the change in LED. The LED current is about 6 mA in both cases.



PM your address if you would like some to try out. I bought quite a few to avoid running out.