On Tue, Jan 14, 2025 at 07:05 AM, Jim VE7RF wrote:
I used an opto-22? ?opto isolator.? You can get em in 5 vdc.....15 vdc...and 24 vdc? on the input side....... (and each of those 3 x different input vdc's is nominal...... IE:? u can go a LOT higher / lower).? ? And they only draw 10 ma on the input side.? ?Output side on any of them is either 60 vdc @ 3 amps? CCS...... or 250 vdc @ 1 amp CCS.? ? Switching time is 100 Usecs for operate, to turn em on.... and 750 Usecs to turn em off, for rls.? ?5 kv isolation between the input and output side.? Input and output sides are isolated via the laser between em...hence the term, opto isolator.??
That's as dead simple as it gets.? ?And NO mech, slow relays involved anywhere. We use em to key the coils of sped up vac relays.? The last thing we want is a 5-15 msec slow relay that in turn keys the coil of a sped up vac relay...which would defeat the initial purpose of the sped up vac relays.? ? Same opto isolator can also be used for cut off biasing on a triode amp. Our friends in VK land have that concept down to a fine artform.? ?The secondary of the opto just shunts the bias cut off resistor.? Again, dead simple.
The Opto-22 has been my go-to method for key-line buffering as well. You can find 'em on eBay (usually brand new) for around $10-$12, although sometimes more. Or, buy them brand-new from Mouser, Newark, etc.
?
I've installed them on my three Henry amps (3K-Classic, 3-KA, 2-KD Classic), as well as in a "rebuild" I did on a 2KD-5 for another ham. Worked great.? Drops the key-line current from >200ma, down to about 13mA.? And as Jim said, they are VERY fast... no relay delay, nearly instantaneous (100 micro-seconds on-time is FAST!).
?
The Opto-22's are tough-as-nails, and in a fairly small package.
?
If all you need is key-line isolation, they're a very easy way to go.
?
?
--
73,
~Alan
|
Re: G3SEK TRIODE BOARD on eBay
WB2FKO said:?
"Keeping with GM3SEK's original design philosophy, the boards could be re-engineered with a modern microcontroller that could also implement a versatile sequencer function, thermal protection, along with surface mount components. I could take a crack at this sometime this spring when I expect to have more time. It would be completely open-source using KiCAD for PCB layout and a Github page for community collaboration.
Mike WB2FKO"
Hey Mike that would be an interesting concept, having a heater timer that could be set for the popular warmup times, including 6 minutes for the YC-156 (do folks still find those tubes anymore?). I started to say there are SMD optoisolators at $.75 each but if I were making it I would use a 6 pin socketed variety so they can be replaced when/if something causes one to fail from excessive current or rf getting into it.? On that subj, be sure to put rf bypassing on I/O pins to keep rf from regurgitating it like the Alpha Radio linears did when they first came out.? I forget the model but it had a row of LED's on the front panel, it tested good into a dummy load at the factory but the first couple of customers found it went bonkers when they used 80M and had rf in the shack so they returned them for Gordon to add rf bypass or ferrite beads or something on the micro. Just a thought but as they say, history seems to repeat itself.
73,? Reid? W6MTF
|
|
|
Jim, Thank you. When studying this, most optocoupler diagrams show 3 connections, VCC, OUT, GND/NEG. For amplifier keying there's only two wires, POS and NEG which need to be connected to key the amp. I wasn't sure how to connect in that manner (three connections). My drawing above just?connected the two wires from?the amp, energising the amp TR relay. How would I have used the three connections? What am I missing? Bob W4JFA
PS. BTW 2N2222 transistors can have different pinouts so always check these. Lots of reading about that on it online.
I used an opto-22? ?opto isolator.? You can get em in 5 vdc.....15 vdc...and 24 vdc? on the input side....... (and each of those 3 x different input vdc's is nominal...... IE:? u can go a LOT higher / lower).? ? And they only draw 10 ma on the input side.? ?Output side on any of them is either 60 vdc @ 3 amps? CCS...... or 250 vdc @ 1 amp CCS.? ? Switching time is 100 Usecs for operate, to turn em on.... and 750 Usecs to turn em off, for rls.? ?5 kv isolation between the input and output side.? Input and output sides are isolated via the laser between em...hence the term, opto isolator.??
That's as dead simple as it gets.? ?And NO mech, slow relays involved anywhere. We use em to key the coils of sped up vac relays.? The last thing we want is a 5-15 msec slow relay that in turn keys the coil of a sped up vac relay...which would defeat the initial purpose of the sped up vac relays.? ? Same opto isolator can also be used for cut off biasing on a triode amp. Our friends in VK land have that concept down to a fine artform.? ?The secondary of the opto just shunts the bias cut off resistor.? Again, dead simple.
Another option is to use a switch transistor.? $1.85 in the mouser catalog.? Switches in 3.6 u secs. They will handle a helluva lot of voltage and current...and are the fastest thing available.? Issue is...IF they ever failed shorted, the OCV ( open circuit voltage), or in my case the overvoltage used to speed up a vac relay, will then appear across the puny switch transistor typ used in most xcvrs...and kaboom, there goes the puny switch transistor in the xcvr.??
The opto-22 was used in several HB metal triode tube amps.? It was used in the 10 kw? 6m amp.....and it is also used in the 10 kw? 160-15m amp.? The opto keys the Jennings? RJ1A used for the input...and also another RJ1A for the cut off bias....and also 2 x Jennings RJ2B's (contacts in parallel)? for the 10 kw output side.?
Local eng buddy of mine used the same config on his YC-156 amp, except he used +170 vdc? to speed up the 26,5vdc coils.? ?The biggest cap he could install in parallel with the series drop resistor was just .5 uf.? ?That's called a ...."RC hotshot" circuit.? ?In his case, he's a hardcore? QSK CW op.? ?When tested individually on the bench, one of the new ( sped up with 170 vdc plus a .5 uf cap across the drop resistor)? RJ2B's would operate in 2.4 msecs.? The 2nd RJ2B would operate in 2.6 msecs.? ? The cap across the drop resistor will speed up the relays even more, BUT it also forms a? RC time constant on release.? For QSK CW use, that RC time constant has to be kept really low.? ?He could have used a lower overvoltage and a bigger cap across the drop resistor.? When T=0, the discharged cap presents essentially a dead short...and the max overvoltage appears on the coil of the vac relay.? ( you can easily measure that, with a fluke 87 DVM in 1 msec peak mode, tnx scott).? ?Once the cap is charged up, it's game over....since the cap will block DC current....and the now alternate path is through the drop resistor that the cap is in parallel with.??
When going back to RX, the charged up cap (which had the same vdc across it as the series drop resistor, since they are in parallel) now has to discharge through the drop resistor....and that takes time.? It takes a full 3 x RC time constants to discharge any cap? down to 5% of it's original max vdc.? ?For SSB use, ( or semi break in CW) a 5-10 msec? RC time constant is used? (IE: 15- 30 msecs for 3 x RC time constants).? ?For? QSK CW, you want a helluva lot faster than that.? ?Reason is, you want that cap fully discharged? between dots and dashes, so the enhanced speed up action of the vac relay can be maintained.?
On Alpha's? older desk top amps, they used? 36 vdc? and a drop resistor, and a cap across the drop resistor, to speed up the 26.5 vdc? RJ1A they used on the output side.? They ran QSK CW just fine.? ?But the RJ1A / Kilovac HC-1 / Gigavac G1H? are all stupid fast when operated at rated voltage.? ?That 8/6 msec operate time they quote is when the minimum pull in voltage is used.?
|
I used an opto-22? ?opto isolator.? You can get em in 5 vdc.....15 vdc...and 24 vdc? on the input side....... (and each of those 3 x different input vdc's is nominal...... IE:? u can go a LOT higher / lower).? ? And they only draw 10 ma on the input side.? ?Output side on any of them is either 60 vdc @ 3 amps? CCS...... or 250 vdc @ 1 amp CCS.? ? Switching time is 100 Usecs for operate, to turn em on.... and 750 Usecs to turn em off, for rls.? ?5 kv isolation between the input and output side.? Input and output sides are isolated via the laser between em...hence the term, opto isolator.??
That's as dead simple as it gets.? ?And NO mech, slow relays involved anywhere. We use em to key the coils of sped up vac relays.? The last thing we want is a 5-15 msec slow relay that in turn keys the coil of a sped up vac relay...which would defeat the initial purpose of the sped up vac relays.? ? Same opto isolator can also be used for cut off biasing on a triode amp. Our friends in VK land have that concept down to a fine artform.? ?The secondary of the opto just shunts the bias cut off resistor.? Again, dead simple.
Another option is to use a switch transistor.? $1.85 in the mouser catalog.? Switches in 3.6 u secs. They will handle a helluva lot of voltage and current...and are the fastest thing available.? Issue is...IF they ever failed shorted, the OCV ( open circuit voltage), or in my case the overvoltage used to speed up a vac relay, will then appear across the puny switch transistor typ used in most xcvrs...and kaboom, there goes the puny switch transistor in the xcvr.??
The opto-22 was used in several HB metal triode tube amps.? It was used in the 10 kw? 6m amp.....and it is also used in the 10 kw? 160-15m amp.? The opto keys the Jennings? RJ1A used for the input...and also another RJ1A for the cut off bias....and also 2 x Jennings RJ2B's (contacts in parallel)? for the 10 kw output side.?
Local eng buddy of mine used the same config on his YC-156 amp, except he used +170 vdc? to speed up the 26,5vdc coils.? ?The biggest cap he could install in parallel with the series drop resistor was just .5 uf.? ?That's called a ...."RC hotshot" circuit.? ?In his case, he's a hardcore? QSK CW op.? ?When tested individually on the bench, one of the new ( sped up with 170 vdc plus a .5 uf cap across the drop resistor)? RJ2B's would operate in 2.4 msecs.? The 2nd RJ2B would operate in 2.6 msecs.? ? The cap across the drop resistor will speed up the relays even more, BUT it also forms a? RC time constant on release.? For QSK CW use, that RC time constant has to be kept really low.? ?He could have used a lower overvoltage and a bigger cap across the drop resistor.? When T=0, the discharged cap presents essentially a dead short...and the max overvoltage appears on the coil of the vac relay.? ( you can easily measure that, with a fluke 87 DVM in 1 msec peak mode, tnx scott).? ?Once the cap is charged up, it's game over....since the cap will block DC current....and the now alternate path is through the drop resistor that the cap is in parallel with.??
When going back to RX, the charged up cap (which had the same vdc across it as the series drop resistor, since they are in parallel) now has to discharge through the drop resistor....and that takes time.? It takes a full 3 x RC time constants to discharge any cap? down to 5% of it's original max vdc.? ?For SSB use, ( or semi break in CW) a 5-10 msec? RC time constant is used? (IE: 15- 30 msecs for 3 x RC time constants).? ?For? QSK CW, you want a helluva lot faster than that.? ?Reason is, you want that cap fully discharged? between dots and dashes, so the enhanced speed up action of the vac relay can be maintained.?
On Alpha's? older desk top amps, they used? 36 vdc? and a drop resistor, and a cap across the drop resistor, to speed up the 26.5 vdc? RJ1A they used on the output side.? They ran QSK CW just fine.? ?But the RJ1A / Kilovac HC-1 / Gigavac G1H? are all stupid fast when operated at rated voltage.? ?That 8/6 msec operate time they quote is when the minimum pull in voltage is used.?
|
Just wondering Bob, what kind of amplifiers do you have?
|
That's ok. I got it working. These interfaces have a few advantages. It uses the 12 volts from the radio as long as I stay below 10 mA. 12 volts on transmit from the Kenwood's DIN accessories jack. In the case of the 850, the internal amplifier relay can be turned off since it's not used. You also don't have another power cord on the desk.
The base resistor that worked the best was approximately 500 ohms.
I still may change a few things yet. We'll see.
Bob W4JFA?
On Mon, Jan 13, 2025, 10:48 AM Robert Herendeen via <bob.herendeen= [email protected]> wrote: My Fault Bob, I thought you had an HL-2200, that is Rons comment. Sorry for the confusion.
Bob N9NRW
|
Re: G3SEK TRIODE BOARD on eBay
I have a set of tetrode boards on my GS23B amplifier and
understand the design pretty well. Their primary functions are
voltage regulation and screen current trip, which has saved me
countless times over the years.?
Keeping with GM3SEK's original design philosophy, the boards
could be re-engineered with a modern microcontroller that could
also implement a versatile sequencer function, thermal protection,
along with surface mount components. I could take a crack at this
sometime this spring when I expect to have more time. It would be
completely open-source using KiCAD for PCB layout and a Github
page for community collaboration.
Mike WB2FKO
On 1/13/25 11:17, Robin Midgett via
groups.io wrote:
toggle quoted message
Show quoted text
Thanks
for that insight, Reid.
With all
that?being said, the electronic methods for control and
protection of triode and tetrode amplifiers haven't
changed....Ian's circuits, from an electronics standpoint,
are still valid. That being the case, duplicating the
circuits with modern components on new boards is entirely
feasible for someone with PCB design experience. Whether or
not anyone is willing to take up the project is really the
crux of the matter, and support of the product is an option
entirely up to the person doing the work. I'm not willing to
take on that project, but it seems likely to me that
someone?would.
Thanks,
Robin Midgett K4IDC
I contacted a friend of Ian's and he dialed him up and
asked to comment on the subject at hand.
Ian said he basically was unable to help, here's his
reply
?
"There have been a few inquiries about reissuing the
boards, but I'm afraid that is no longer workable. Here's
why:
1. Those boards were designed more than 25 years ago, and
many of the components are now obsolete. This was already
becoming a problem when I closed the business, and of
course has only become worse. Also the software that I
used to create the boards and schematics is way out of
date, so I can't even edit the files any more.
2. Even if by some miracle I could find the Gerber files,
all of those other changes mean that re-issued boards
would in reality be new products, each with its own new
range of support issues. I always took the "lifetime
support" commitment very seriously, but today I don't even
have the amplifiers or the test rigs to simulate faults or
provide support. Too many things have changed."
?
That's life, things change.
73
Reid Brandon? W6MTF
?
|
Re: G3SEK TRIODE BOARD on eBay
Thanks for that insight, Reid. With all that?being said, the electronic methods for control and protection of triode and tetrode amplifiers haven't changed....Ian's circuits, from an electronics standpoint, are still valid. That being the case, duplicating the circuits with modern components on new boards is entirely feasible for someone with PCB design experience. Whether or not anyone is willing to take up the project is really the crux of the matter, and support of the product is an option entirely up to the person doing the work. I'm not willing to take on that project, but it seems likely to me that someone?would.
Thanks,
Robin Midgett K4IDC
I contacted a friend of Ian's and he dialed him up and asked to comment on the subject at hand.
Ian said he basically was unable to help, here's his reply
?
"There have been a few inquiries about reissuing the boards, but I'm afraid that is no longer workable. Here's why:
1. Those boards were designed more than 25 years ago, and many of the components are now obsolete. This was already becoming a problem when I closed the business, and of course has only become worse. Also the software that I used to create the boards and schematics is way out of date, so I can't even edit the files any more.
2. Even if by some miracle I could find the Gerber files, all of those other changes mean that re-issued boards would in reality be new products, each with its own new range of support issues. I always took the "lifetime support" commitment very seriously, but today I don't even have the amplifiers or the test rigs to simulate faults or provide support. Too many things have changed."
?
That's life, things change.
73
Reid Brandon? W6MTF
?
|
Re: G3SEK TRIODE BOARD on eBay
I was reminded there is another source for boards to use with a triode OR tetrode, one built in France
A couple of years ago I was in a discussion about using one of them on a larger tetrode in Europe.
Not certain how that turned out but I bought one of the boards but never completed the project.
Check out the link, it may come up in French but there is a button to translate to English.
?
Dom is an active designer and quite friendly and able to communicate despite the language barrier.
This should then be considered a new source of boards for use with tube linears.?
73,
Reid W6MTF
|
Re: G3SEK TRIODE BOARD on eBay
I contacted a friend of Ian's and he dialed him up and asked to comment on the subject at hand.
Ian said he basically was unable to help, here's his reply
?
"There have been a few inquiries about reissuing the boards, but I'm afraid that is no longer workable. Here's why:
1. Those boards were designed more than 25 years ago, and many of the components are now obsolete. This was already becoming a problem when I closed the business, and of course has only become worse. Also the software that I used to create the boards and schematics is way out of date, so I can't even edit the files any more.
2. Even if by some miracle I could find the Gerber files, all of those other changes mean that re-issued boards would in reality be new products, each with its own new range of support issues. I always took the "lifetime support" commitment very seriously, but today I don't even have the amplifiers or the test rigs to simulate faults or provide support. Too many things have changed."
?
That's life, things change.
73
Reid Brandon? W6MTF
?
|
My Fault Bob, I thought you had an HL-2200, that is Rons comment. Sorry for the confusion.
Bob N9NRW
|
You could just use the sb-220 softkey circuit and eliminate the need for the external power.?
I stuck it on a strip board.?
?
|
It looks like my drawing will work. I used a 2N2222 to switch it. This is done on a perf board only. The amp keying is 25 volts?@ 60 mA so I'll have to determine the 2N2222 base resistor. I'm measuring just under 7mA on the diode side which is what the radio will see. I would like it down to 5 or 6mA.?
To get the voltage and current down on the diode side I made a voltage divider?which brings the voltage down to around 6 1/2, which is where I'm measuring?the diode current at just under 7mA.?
Still working at it.
Bob W4JFA
Ron, Thank you. I'm trying to learn these optocouplers?so I need to figure out on my own. Attached is my drawing. After looking around I may order? a few 4N25's. Bob W4JFA
Another simple almost complete interface costing a few $$:? An Arduino 12V relay module.? It has an optoisolator input driving a reed relay.? 12V power, and a couple cables are all that's needed to make the interface.? I've used it to connect my Ten Tec Jupiter to my Heathkit HL2200 linear.??
?
?
|
Ron, Thank you. I'm trying to learn these optocouplers?so I need to figure out on my own. Attached is my drawing. After looking around I may order? a few 4N25's. Bob W4JFA
Another simple almost complete interface costing a few $$:? An Arduino 12V relay module.? It has an optoisolator input driving a reed relay.? 12V power, and a couple cables are all that's needed to make the interface.? I've used it to connect my Ten Tec Jupiter to my Heathkit HL2200 linear.??
?
?
|
Years ago Gary made a relay-driven protector
when it was seen that a large RF signal could
damage the circuitry of the transceiver’s RX2
port.
?
It turned out to be an excellent RF amplifier
keying driver as the high quality of the driver
relay provided a worthwhile isolation.
?
There are a number of available products
these days, I hear. Like ?I.C.E. Products,
the KD9SV stamp is to investigate.
?
Hal
W4HBM
?
?
|
Another simple almost complete interface costing a few $$:? An Arduino 12V relay module.? It has an optoisolator input driving a reed relay.? 12V power, and a couple cables are all that's needed to make the interface.? I've used it to connect my Ten Tec Jupiter to my Heathkit HL2200 linear.??
?
?
|
Hi all,
Since I'm now retired I have time to goof off. I was trying to design an amplifier keying interface using an optocoupler. I don't need one since I already?built a few using BJT's?and/or darlingtons so this is just killing time on a rainy day.
I have an MCT61. I haven't been able to figure out the datasheet since it seems to list things in a different way. The?datasheet is probably fine, it's just I have virtually no experience with optocouplers.
So using the DATA DIN on the back of my 850 and 870 I need to keep the current draw on that DIN under 10 mA so my target is 6 mA.?
I made a voltage divider?to keep the voltage down to the opto and also keep the current down on the radio's 12 volt supply. I did hit my target of 6 mA and also I measured 6 mA on the diode side of the opto but on the transistor side of the opto? I measured 85 ohms from C to E when turned on so I feel the?opto was not completely turning on the transistor side. I was hoping to see zero ohms when turned on, not 85 or so.
What I was going to do next is use a regular switching BJT driven by the transistor side of the opto to do the amplifier keying/switching. I haven't got to that part of this experiment yet but my gut feeling is I don't have the correct optocoupler for this project. Again I need to keep the current draw on the Kenwood radio's DIN plug at 6mA so my thinking is I need an opto that will completely saturate the transistor side with 6mA current on the diode side and it doesn't look like what I have can do that.
Again, this is just for fun but if I'm successful?I'll make one for my amps.
Am I on the right track?
Thanks for reading.
Bob W4JFA
|
Re: G3SEK TRIODE BOARD on eBay
The Art Work for the PCBs is the property G3SEK. The circuitry is nothing special and is Public Domain. Using any one of the popular PCB Design CAD Programs (Altium Designer/Circuit Studio/Maker), Novram Diptrace, or Kicad those boards could be designed and the Gerber Files generated with a weeks worth of work.
Hullo Tracey and the Group,
?
Does anyone chat about with Ian, GM3SEK?
?
He’s a member of the Oxford University Radio
club, and still active on VHF, I believe.
?
Perhaps he might “Pass along the ball” to
the next generation, seeing as how his
design has been growing in popularity.
?
Hal Mandel
W4HBM
|
Re: G3SEK TRIODE BOARD on eBay
Hullo Tracey and the Group,
?
Does anyone chat about with Ian, GM3SEK?
?
He’s a member of the Oxford University Radio
club, and still active on VHF, I believe.
?
Perhaps he might “Pass along the ball” to
the next generation, seeing as how his
design has been growing in popularity.
?
Hal Mandel
W4HBM
|
Alan - W5ARM
Brandon DX
Bob
Robin Midgett
HaL Mandel