- Amateur Radio (Ham Radio) Community

Call Search

New to Ham Radio?
My Profile

Friends Remembered
Survey Question

DX Cluster Spots

Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement

[Articles Home]  [Add Article]  

Henry 2K-4 QSK Modification

Thomas A. Boza (NE7X) on February 4, 2011
View comments about this article!

Henry 2K-4 RF Amplifier QSK Modification


My ICOM IC-7700 transceiver was showing high infinity SWR for approximately 25ms every time I keyed the transmitter, before it would go to 1:1. This happened only when the Henry Radio 2K-4 RF amplifier was in-line.


When the IC-7700 PTT is keyed, RF output does not occur until after 6ms. This 6ms RF output delay it designed into the IC-7700 to allow any external T/R switching to complete before RF output. However the internal mechanical T/R relay in the Henry 2K-4 was taking ~31ms to fully energize and complete switching state. This caused the IC-7700 to transmit into an open circuit for ~25ms.


Transmitting into an open circuit for ~25ms every time I pressed PTT could cause the RF output transistors in the IC-7700 to go into oscillation and possible damage to occur.

NOTE: This issue also applies to most modern transceivers, not just the IC-7700.


Checking with ICOM technical support, there are no user or factory hidden service menus for the IC-7700 to adjust or extend the 6ms RF output delay. This delay is a fixed static value. The IC-7700 also DOES NOT have a transmit inhibit input for closed loop.


To resolve this problem I decided to replace the slower mechanical T/R relay in the Henry 2K-4 with high speed Gigavac model GH1-HAM (datasheet) 5KW RF 6ms 12VDC SPDT vacuum relays. Not only will this resolve the issue of having the IC-7700 transmitting into an open circuit every time I key the transmitter, it will also allow the amplifier to run full QSK break-in. I found the GH1 vacuum relays actually switched in 3ms. The 6ms is the manufacture's worse case value.


This modification requires three SPDT GH1 vacuum relays. One for switching the Henry 2K-4 antenna RF output, one for switching RF input, and one to pull the 3-500Z grid bias to ground.

Since this modification utilizes three separate SPDT vacuum relays, the IC-7700 and the RF amplifier's vacuum relays MUST all be switched in a certain sequence. Sequencing is required to prevent any RF damage from occurring to either the RF amplifier output circuit (3-500Z tubes), vacuum relay contacts (arcing and pitting) or the IC-7700 transceiver. The sequence is:

The IC-7700 PTT is keyed

  1. RF OUT vacuum relay in the Henry 2K-4 is switched
  2. RF IN vacuum relay in the Henry 2K-4 is switched
  3. BIAS vacuum relay in the Henry 2K-4 is switched
  4. IC-7700 RF output occurs

This sequence is equally important when un-keying from transmit, The IC-7700 must stop RF output first and the three vacuum relays must be un-energized in the reverse order as they were energized.

To accomplish sequencing for the vacuum relays, I added a JWM Engineering Group  four step port programmable SEQ-QSK T/R Sequencer.


SEQ-QSK Specifications

  • 13.8 Vdc @ 50ma typical (no switched output active)
  • 35 Vdc @ 600 millamperes open collector switching capability, Four outputs total
  • Enable High or Enable Low input to activate sequencer
  • Visual states displayed on 4 Green State LED's
  • 1 second Yellow Heartbeat indicator
  • Programmable delay times from 1 ms to 32 ms (default is 15 ms)
  • Operating Temperature Range 0 Deg C to +70 Deg C
Complete step-by-step details on how to perform the QSK modification can be found at the following web link:

Member Comments:
This article has expired. No more comments may be added.
Henry 2K-4 QSK Modification  
by AD5X on February 4, 2011 Mail this to a friend!
I don't know about the IC-7700, but the IC-7600 can have a QSK turn-off problem. The IC-7600 unkeys the amp 5-7ms before RF goes away. So if your QSK relays or PIN diodes unkey faster than this, you will hot-switch. I actually have a 'scope trace provided by Icom that shows this. See

The unkey issue is also a problem operating QSK with the IC-706MKIIG. In my above referenced article, I measured my IC-706MKIIG timing in QSK mode. And finally, there was a recent QEX article that documented this same issue when using an OmniVI TX OUT/TX EN output for driving a QSK amplifier.

Phil - AD5X
RE: Henry 2K-4 QSK Modification  
by N3JBH on February 4, 2011 Mail this to a friend!
Thanks Phil i was not aware there was a issue with the OmniVI Jeff
Henry 2K-4 QSK Modification  
by KASSY on February 4, 2011 Mail this to a friend!
The IC-7700 amplifier can oscillate when fed into an open circuit? I'm not an EE, but took a few classes, and I think that's called "conditionally unstable". I'm pretty sure that nobody would really release an amplifier to the market that would do that. They don't like open circuits, but they should not oscillate. Maybe an EE onsite can I off base?

- k
RE: Henry 2K-4 QSK Modification  
by NB3O on February 4, 2011 Mail this to a friend!
"can oscillate"
I have a few IC-746's which will oscillate into anything greater than 3:1 VSWR when using them on 2 meters.

I have not had the cojones to load-pull their HF amp side, since I blew out a Kenwood TS-440 final trying to benchmark its' behavior with the old triple slide trombone.

I've come to the conclusion most of the current rig manufacturers do not do a transmitter load-pull into anywhere near an open (around Mr. Smith's chart perimeter), nor do they guarantee unconditional stability.
That's why a lot of them come with internal tuners and have temperamental VSWR fold-back circuitry.

But then, they are more affordable than the commercial stuff.

RE: Henry 2K-4 QSK Modification  
by W8NF on February 4, 2011 Mail this to a friend!
I doubt if oscillations killed that TS-440 amp. I don't know what devices it used, but about that time, most manufacturers transitioned from 2SC2290 to 2SC2879. The general design philosophy with the '2879 is to design the amp for about 200 watts Psat, then drive it gently enough to get half that. Efficiency suffers, but at 100W, that amp cannot be blown by SWR - no SWR at any phase angle will develop enough peak collector voltage or emitter current to cause a problem. However, if you get a guy who modded the rig for "more power", then all bets are off. Also, Kenwoods love to overshoot due to their sluggish ALC systems, and while in overshoot, you can kill them.

If the rig used the 2SC2290, however, they were indeed rather marginal, and many a pair died due to instantaneous overvoltage during a high SWR condition. Not oscillations.

Looking back into the amp with a VNA and making sure that there is no frequency at which they exhibit negative resistance should do it...the exception is if there's something lossy between the output and the devices, that's preventing the VNA from seeing the true Zout (when not driven, but biased). "Some bets are off" in that case anyway, because when biased but not driven, you're not swinging the device through all of its I-V region.

The IC7700 uses a pair of MRF150s, or so I've been told. I did design many an MRI amplifier with these devices, and with reasonable feedback networks (we used 25 ohms in series with about 0.1uF...there would be less power loss with transformer-coupled feedback), they are, indeed, unconditionally stable. We verified same with both a VNA and a load pull. Of course, Icom may well have used a differnt topology than we did.

The built-in tuners are, IMO, a deficit vis-a-vis stability. They can present some pretty crazy impedances to the output devices at harmonic and other out-of-band frequencies. I would not dare put a T tuner on an amp until I had confirmed its unconditional stability.

Not knowing the OP, I am speculating, but I suspect the claim of oscillation was simply there because he'd read it about lots of rf amplifiers and figured they were all capable of it. ***sigh*** Sadly, probably just another example of a cleverly-written web page with very poor technical advice being quoted by so many people that the community at large started believing the non-proofs. A classic case of a "meme" if there ever was one!


Dave W8NF
-who just took notice of the VERY nice LDMOS devices coming from NXP and Freescale and is getting the "I'm tired of microwaves, it's time to do power again" urge.
Ye olde relay trick  
by N9NFB on February 6, 2011 Mail this to a friend!
Absolutely nothing wrong with upgrading an amp, nor making T/R switching faster, for very practical reasons.

However. The stated hot switching safety problem can be fixed by ye olde two relay trick. Obtain a male and female microphone plug/socket. Straight-thru wiring between them. Operate for awhile to verify. Then modify by sampling the ground and breaking the PTT.

Make one relay key "instantly" off the mic PTT and the other take, per your article, about 40 ms longer. Perhaps you can use one pair of contacts on the first relay to control the coil on the second relay. Perhaps the first relay gets a simple one transistor "instant" keying circuit and the other gets a 555 timer built for about 40 ms. Perhaps you have two simple 555 timers for "debouncing" the contacts one takes 40 ms longer than the other. Maybe as simple as two rectifier diodes and a big-ish electrolytic cap would be adequate. Or you can go all microcontrollery, an 8pin PIC10F220 and about ten lines of code should be more than adequate to the task. You get the idea, the point is when you hit PTT, one DPDT relay keys 40 ms before the other relay, and when you unkey, the slow one one also hangs on 40 ms later.

Now wire up the following logic functions off the relay contacts. slow AND fast = radio, slow OR fast = amp. Or rephrased the amp contacts are in parallel and the radio contacts are in series. Or rephrased, the amp keys on at "fast relay" speed and off at "slow relay" speed, whereas the radio keys on at "slow relay" speed and off at "fast relay" speed. Hot switching seems physically impossible with this wiring, until a relay coil fails or a contact welds. If you buy a more than 2 throw relay, you can wire up a little protection circuit that prevents operation if either relay fails or welds shut (or for that matter welds open)

Relay logic gets kinda complicated when you're doing the full microwave experimenter thing with a xverter and a preamp so at that level you pretty much need the full sequencer, but ye olde dual relay trick works adequate for just a radio and an amp.

Just to reiterate this is a perfectly valid performance upgrade in and of itself, but is a bit elaborate to merely prevent hot switching.

73 de Vince N9NFB EN53ua
RE: Ye olde relay trick  
by WB2WIK on February 6, 2011 Mail this to a friend!
Nice article.

I like that my Ten Tec rigs, even including the little Jupiter, have all of these timings fully adjustable, to accommodate any kind of amp.

I can set the rig to key the amp 50 mS before any RF output occurs, so this problem would never occur.

Since all this can be done in software, I have no clue why Icom or anyone else wouldn't enable a similar feature in all their rigs.
RE: Ye olde relay trick  
by WA2JJH on February 13, 2011 Mail this to a friend!
Steve, possible reason why all the new rigs do not have what the Jupiter does have.

It is similar to why the old Heathkit SB-220A/HL-2200 have to have many mods to be added for use with even a TS-850.

The input filter network you and everybody knows about.

Could it be a money matter?

JA made rig corps do not want folk to modify perfectly good Heathkit gallon out+ amps.
They are in a business. They want Hams to spend top dollar on new JA made amps.
The Kenwood twin 3-500 amp is inferior to an older
Heathkit twin GG 3-500z amp as all know.

They have a vested interest. They want us all to buy solid state amps like the QUADRO.

You should have seen the damage an old SB-220A/HL-2200 did to one of my TS-850's. I did wise up and rebuilt the HL-2200.
It is all about the money!!!!! well maybe.
RE: Henry 2K-4 QSK Modification  
by W9AC on March 4, 2011 Mail this to a friend!
> "Thanks Phil i was not aware there was a issue with the OmniVI Jeff"

The use of combining the Omni Six TX-OUT and TX-EN lines with a "Y" adapter results is hot switching. I had documented this back in 1998. Search the Ten Tec list on for more information.

A better way to use the Omni Six for QSK operation into high-speed amplifiers is almost as simple: Cut the relay output key line and bring the switching transistor's collector out to the key line jack. The transistor's base resistor is then tied to the "T" line voltage buss.

Paul, W9AC
Email Subscription
You are not subscribed to discussions on this article.

My Subscriptions
Subscriptions Help

Other How To Articles
How to Use a Spectrum Analyzer