Transmitter Biasing for Optimum Modulation & Stability

Discussion in 'Technical' started by W5HRO, Apr 8, 2014.

  1. W5HRO

    W5HRO Administrator

    Attached below is my HB transmitter biasing and negative peak limiter circuit. The only mods I have left to do are to find a better 10H higher voltage choke (1000V min) for the screen circuit then modify the fixed control grid bias supply to where it's -200V on idle and then switches down to -100V during transmit for safety biasing. Both the control grid and screen supplies are regulated with VR tubes. The control grid supply mod will be the tricky one.

    I think I can easily make the control grid supply switch from regulated -200V to -100V if I just tap into my PTT control circuit. The trick will be in configuring everything and making it switch between the voltages without using a mechanical relay if possible. I want it to be all electronic. I may need to use an SCR or two. The control grid supply was designed for two 4-400's so I think a voltage divider setup should work because the supply will still have enough current. Maybe even just tapping it between the two series VR tubes for the lower voltage would work as well if not better.

    The whole design idea is to have -200V on the control grid during idle like when spotting signals via the receiver and not have the RF exciter's load shift and possibly change frequency like it might if it was always at -100V instead then gets blocked by D4 and D5 when it increases more negative during transmit with RF drive applied which in turn also provides 100% grid leak bias for optimum control grid modulation.

    P.S. The 4-400 is a PL 4-400CG designed for +500V on the screen and -200V on the control grid.

    Transmitter_Bias.png
     
  2. W5HRO

    W5HRO Administrator

    I think using two 0B2's would be perfect. The PL-4-000CG is designed for 12mA control grid bias at -200V with a plate voltage of +2000V and -220V with a plate voltage of +2500V. So, with my +2125V plate supply around -210V to -208V is just right and then split that in half for the real time safety bias. I also already have 3 high wattage non-inductive 18K resistors. I can use two in series to get the 36K for R2 and the other one for the grid leak resistor, R3.

    Anyway, after considering that along with other things is the updated circuit below. My bias supply deck already has the two VR tubes is series so this will be an easy modification. I just need to find the right SCR in a TO-3 package to mount on the back of the chassis, but it will need to be isolated from ground since it's a negative supply. That still depends on the type of SCR though. Even a solid state relay brick would probably work. There are no mechanical relays in the transmitter and I want to keep it that way. I hate mechanical relays about as much as mercury vapor rectifier tubes.

    The paralleled 1N4007's are to split the currents and just make everything more robust.

    Control_Grid_Supply_Mod_02.png
     
  3. W5HRO

    W5HRO Administrator

    I had some time last night so I quickly built up the circuit in my last post using clip leads and a SS relay brick for the SCR which is an SCR anyway, but controlled by 110Vac. I tapped over to my screen supply transformer's primary which gets powered up with PTT to control the SCR. My grid and screen transformers are Desk KW transformers, same as in the KW-1. What I need to do though is make the SCR on during idle then open at PTT which is the reverse.

    Anyway, it actually worked great and I can switch the grid supply voltage up and down with the PTT and the 0B2's stay stable either way. I'll post a final circuit if I need to make any changes.
     
  4. WD5JKO

    WD5JKO Member

    Brian,

    It has been a long time since I've played with SCR's or even a triac. Therefore I cannot say much about your circuit other then I don't understand it. A SCR as I recall is a PNPN device, and to turn one off after it is ON, you must either lower the forward current below a minimum holding current, or reverse bias the device. This gets easy when the supply voltage is AC. I once dealt with self commutation SCR DC to AC inverters at the 10KW level...Boy those would twitch 1 ought wire when the commutation failed to occur. These were fed with large lead acid storage batteries adding up to about 60 volts. We got tired of blowing the big fuse, so it got bypassed...When the inverter lost commutation again of one of the SCR's we had a "China Syndrome" emergency event on our hands...

    Anyway I hope your circuit works as intended, and I will wait to see how the finished product works out.

    Jim
    Wd5JKO
     
  5. W5HRO

    W5HRO Administrator

    An SCR is just a latch, like opening and closing relay contacts. When latched current flows through the diode.

    Anyway, I'm not sure what you don't understand though. The SCR should be latched at idle so that the full grid supply is around -208V. Then during transmit the latch opens and then only around -103V is applied to the grid to act as a real-time safety bias during grid leak biasing via the 18K resistor.

    When I hooked up the temporary test circuit the other night I was using a SS relay, but it was an AC controlled relay so I had the reverse SCR condition. The main thing I wanted to see is if the supply would switch up and down between the 2 voltages and it did. I can either use a DC controlled SS relay (also an SCR) or a stud mounted SCR. There's a couple of different ways to latch then unlatch a stud mounted SCR. The trick is making them unlatch again after they are already latched. The SS relay bricks just have all of that built in.

    I still have several DC controlled SS relay bricks so I may just use one of those and leave the control DC on at idle then have it open (go off) during transmit. The SCR in my circuit can either be a stud mounted SCR or just a DC controlled SS relay, but the SCR symbol is there in the drawing for simplification either way.

    Maybe what you are missing is the fact my grid bias supply is already there with the two 0B2's and I'm adding this simple modification so I can use grid leak bias during transmit. I just want the grid voltage to jump back up to the full -208V each time where it currently is anyway during idle.
     
  6. W5HRO

    W5HRO Administrator

    You know, a GTO (Gate-Turn-off-Thyristor) would actually be best because they are easy to turn back off.

    I found the ones below, but I don't know what the max voltages are. Anyway, that would be the alternative if they work with at least 400Vdc or 500Vdc. I think they are probably very high current up to like 1600V. They could maybe even be used to turn the screen supply on and off as long as they were properly bypassed on both sides and at the control input.

    http://cnrectifiers.en.made-in-chin...na-Gate-Turn-off-Thyristor-GTO-Thyristor.html

    Control_Grid_Supply_Mod_03.png
     
  7. WD5JKO

    WD5JKO Member

    I've never played with a GTO, although I've ridden in one. :icon_shifty:

    While searching around for parts, I came across a neat series of IXYS FET output opto-couplers.

    I'm not sure this link will be lasting, but here is the line sold by Digikey:

    http://www.digikey.com/product-sear...t=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25

    If the link goes bad, search in Digi-Key for:

    Solid state relay

    SPST

    IXYS manufacturer

    CPC Optomos

    ISOPLUS264 Package

    DC

    SPST

    Normally Open


    I attach the data sheet here of the CPC1788J. This is a Opt-FET where the FET can withstand 1KV off, and pass 1A ON. The FET RDS On resistance is only 1.25 ohm.

    The applications here to vinatage BA rigs are many. Some items come to mind:

    Key the B+ to the final and modulator together or separately in 100w class rigs like the Vikings, Rangers, DX-100, etc. Here you can keep the HV supply ON, and just key the B+. This would make T/R time sequencing easier.

    Key the B+ to a receiver as a mute means. Perhaps just drop the audio output tube B+, or IF amp B+ so that the oscillators keep going while receiver is muted.

    Lots of stuff...

    Jim
    Wd5JKO
     

    Attached Files:

  8. W5HRO

    W5HRO Administrator

    I'm probably just going to use a 10 to 20A solid state relay brick. 1A isn't going to cut it for me, because I always overbuild everything on purpose for reliability. The trick is going to be controlling it in the reverse condition. What I mean is having it normally turned on to where the control voltage goes back to zero during xmit. I want to use my simple PTT control so its not going to be that easy unless I do something like use a pull-up resistor on the positive control terminal to where it pulls down to ground during xmit. That might actually be doable.

    The other thing is modifying my regulated screen supply to where the primary is always on then just breaking the 500V at idle. I might be able to use the same source control supply for an additional solid state relay brick.

    Something like the below might work.

    SS_Relay_Control_1.png
     
  9. WD5JKO

    WD5JKO Member

    OK, so when you close the PTT, the screen turns on, and when PTT opens the screen is turned off, and the cutoff grid bias is applied?

    Oh, that IXYS solid state relay I mentioned is rated at 2.5 amps CCS with a small heatsink, and 1 amp CCS without. Looking at the last curve in the data sheet, it shows it is good for no more than 12 amps with a limit of 10 ms. So keying a capacitive load needs a series resistor to keep the charging current under 12 amps peak. So for a 600v supply, a 50 ohm series resistor will limit the peak charging current to 12 amps. I was thinking of a case such as the Viking II where two of these could be used; one for the RF, and one for the modulator with staged keying such that the modulator comes on 100ms AFTER the RF, and then off 100ms before the RF is turned off. Kind of cool that these parts are available that we can play with.

    I wonder if you have a particular type of Solid State Relay picked for your application?

    Jim
    Wd5JKO
     
  10. W5HRO

    W5HRO Administrator

    Yes

    I have both 240V and 480V ones out in the garage, but I don't have one for the screen supply yet. They are normally rated in AC voltage, but I think most of those ratings are in RMS which should also be good for DC.

    Attached below is a link for some others and they have one for 530Vac, but I don't think that's high enough. I would prefer a 1000V one or close to it and I have seen them before.

    http://sicoelectric.en.alibaba.com/product/706129765-213306934/SSR_solid_stater_relay.html

    The ones you posted would be good for use in receivers for muting, but I have a different opinion as to their use in the old transmitters you listed. I really don't think its a good idea to leave the plate supplies turned on in those things because the transformers are too old. Plus the safety risk involved by leaving them on. They would probably crap out in a short amount of time after being left on and somebody might get electrocuted.

    P.S. The old plate transformer in the DX-100 / TX-1 is the worst one. I had about 3 of those things crap out thru the years. Two in DX-100's and one in a TX-1.

    Anyway, those IXYS relays would need the heatsinks that slide on and sort of clamp down via spring type action because there are no mounting holes for screws. Those things are designed to be soldered down on a PC board from the backside and you cannot do that by hand. I think even the GTO I posted might have PGA or LGA pad type connections on the backside and may be an issue.
     
  11. W5HRO

    W5HRO Administrator

    Here's one that might work for a +500V screen supply (HS501DR-HD6025). It's good for 660Vrms with a 1200V transient overvoltage rating.

    I've even seen them with a little higher voltage than this before, but you really have to search around to find them. The one attached below would probably work fine for my supply and long as all 4 terminals were well bypassed/filtered.
     

    Attached Files:

  12. W5HRO

    W5HRO Administrator

    Oh, attached below is a block diagram of what solid state relays look like in general. They are basically an SCR with all of the control circuity built in.

    Block_Diagram.png
     
  13. WD5JKO

    WD5JKO Member

    Brian,

    The image you show is correct for DC, an optically coupled SSR for DC which has a FET output. The device you attached in the prior post as a PDF is for AC (as stated on datasheet) and has either dual SCR's or a single triac in it.

    Here is a discussion about using an AC rated SSR on DC:

    http://bbs.homeshopmachinist.net/threads/50917-Solid-state-relay-question

    The Crydom folks have a good reputation with their line of SSR's. That said, they are quite stingy on detail of what they put inside their modules. Here is a link from them concerning DC SSR applications, and the need for a diode with inductive loads:

    http://www.crydom.com/en/Tech/Tips/DC relays-technical guidelines.pdf

    Jim
    Wd5JKO
     
  14. W5HRO

    W5HRO Administrator

    Most all of those relay outputs are for AC, but many will work fine with DC too. They just short the drain and source when enabled which will pass anything, AC or DC. If they have triacs then that is a different matter entirely.

    P.S. A lot of the ones too that won't work on DC are capacitor coupled to one of the terminals. Some of them are like that so you have to be careful. The orange ones in my HB transmitter are like that, they measure around .05uF across the output terminals when turned on.
     
  15. W5HRO

    W5HRO Administrator

    By the way attached below are the ones for 1000Vdc. This is what I would prefer to have for the screen supply, but the minimum price for just one is around $200. That's the problem, they are very expensive.
     

    Attached Files:

  16. W5HRO

    W5HRO Administrator

    Damm… these 1000V 25A Crydom relays are heavy :mrgreen: I found a surplus source.

    1000V_25A.jpg