SV 811-10, Again ( I'm a bad poet, and I know it )
The SV 811-10 continues to thwart me on the modified SSE board, but, as a consolation prize, some meaningful progress has been made on the TSE board.
When last seen, it had darn near set my board on fire from the crazy gas on stupid fire, cavalier attitude, of trying to extract almost nine amps from three amp parts.
In place of the 1N5401 diodes, I installed Vishay MBR1645 Schottkey (sp?) rectifier diodes, Mouser p/n 625-MBR-1645-E3, which are nominally rated at 16 amps, 45 volts, and a maximum 150C junction temperature at full current.
These have to be heatsinked, and there is scant real estate in this area of the board for that. I used the smallest TO-220 heat sinks I had on hand, and still had to fiddle with the fins to squeeze them into this tight area. ( Note to self - install the heat sinks BEFORE you mount the parts ) Installation was further complicated by the damage in this area of the PCB where all of the traces were scorched beyond use from round one with this tube, and have to be rebuilt on the back of the board.
I am guardedly optimistic that these present a workable solution. After three hours of continuous operation, which should be sufficient to reach a full heat soak, I recorded the following temperatures at 72 F ambient air temp:
Power Transformer - 117 F on lamination stack ( won't even warm coffee )
Dual Schottkey - 195 F
Schottkey #1 - 212 F ( middle diode )
Schottkey #2 - 198 F
Hot, but well within the ratings.
These diodes produce 7.5 volts dc at the filaments of the 801A, and 6.3 volts dc on the SV-811. Right on the money.
With the SV-811, I used a Sylvania fat boy 5AR4, thoughtfully mismarked for the government as a 5V4GA. Loaded B+ is 445 volts. I did not try to exceed 450 volts, so as not to risk smoking another 10M45.
The Hammond 125ESE were set to 5K. Bias was set to 67 ma. ( My two tubes are not identical - one would not bias beyond 67 ma, therefore these numbers correspond to the colder tube )
The two tone, 6 kHz, 60 Hz, 4:1 numbers:
3 watts - 13% IMD
2 watts - 8.5 % IMD
1 watt - 5% IMD
500 mw - 4.6% IMD
250 mw - 3.2% IMD
100 mw - 3 % IMD
Clean 1 kHz sine wave output to an 8 ohm non inductive load was 9.5 watts. The idle power was 30 watts, representing almost 33% efficiency, which seems remarkable.
I tried 2.5K plate load, and the tubes didn't like that. I tried 10K and got numbers not materially different from what I have posted for the 801A. Clean sine wave output was 6.125 watts at 10K.
Subjectively, this tube is the real deal. The speakers in my ham shack are some little Yamaha monitors. The room is 8 x 12. Where the 801A begins to break up on bass, the SV 811 hunkers down and shakes the walls. Literally. This is a 65 watt tube, and has at least another 2 to 3 dB left in it, and you can really hear it.
I see no reason why 15 to 20 watts SE cannot be realized from this tube, on this board.
I see several ways to get there. SV-811-3 would be more compatible with the unmodified bias circuit and would almost certainly get there, but might present drive problems with the low mu. I don't have any, and they are expensive, so I am unlikely to acquire any.
The bias circuit could be modified to present a lower bias voltage, but this might make the circuit incompatible with the 801A. This might be viable.
The B+ could be voltage doubled and then adjusted, but again this might make the circuit incompatible with 801A which has a 600 volt limit, IIRC. This also might be viable.
An external custom power supply may be the best option and would also be useful for crushing 6146's.
With so many projects in the pipeline, I may just enjoy this modest amount of success before I come back to this tube on this board.
Happy New Year to all.
Win W5JAG
The SV 811-10 continues to thwart me on the modified SSE board, but, as a consolation prize, some meaningful progress has been made on the TSE board.
When last seen, it had darn near set my board on fire from the crazy gas on stupid fire, cavalier attitude, of trying to extract almost nine amps from three amp parts.
In place of the 1N5401 diodes, I installed Vishay MBR1645 Schottkey (sp?) rectifier diodes, Mouser p/n 625-MBR-1645-E3, which are nominally rated at 16 amps, 45 volts, and a maximum 150C junction temperature at full current.
These have to be heatsinked, and there is scant real estate in this area of the board for that. I used the smallest TO-220 heat sinks I had on hand, and still had to fiddle with the fins to squeeze them into this tight area. ( Note to self - install the heat sinks BEFORE you mount the parts ) Installation was further complicated by the damage in this area of the PCB where all of the traces were scorched beyond use from round one with this tube, and have to be rebuilt on the back of the board.
I am guardedly optimistic that these present a workable solution. After three hours of continuous operation, which should be sufficient to reach a full heat soak, I recorded the following temperatures at 72 F ambient air temp:
Power Transformer - 117 F on lamination stack ( won't even warm coffee )
Dual Schottkey - 195 F
Schottkey #1 - 212 F ( middle diode )
Schottkey #2 - 198 F
Hot, but well within the ratings.
These diodes produce 7.5 volts dc at the filaments of the 801A, and 6.3 volts dc on the SV-811. Right on the money.
With the SV-811, I used a Sylvania fat boy 5AR4, thoughtfully mismarked for the government as a 5V4GA. Loaded B+ is 445 volts. I did not try to exceed 450 volts, so as not to risk smoking another 10M45.
The Hammond 125ESE were set to 5K. Bias was set to 67 ma. ( My two tubes are not identical - one would not bias beyond 67 ma, therefore these numbers correspond to the colder tube )
The two tone, 6 kHz, 60 Hz, 4:1 numbers:
3 watts - 13% IMD
2 watts - 8.5 % IMD
1 watt - 5% IMD
500 mw - 4.6% IMD
250 mw - 3.2% IMD
100 mw - 3 % IMD
Clean 1 kHz sine wave output to an 8 ohm non inductive load was 9.5 watts. The idle power was 30 watts, representing almost 33% efficiency, which seems remarkable.
I tried 2.5K plate load, and the tubes didn't like that. I tried 10K and got numbers not materially different from what I have posted for the 801A. Clean sine wave output was 6.125 watts at 10K.
Subjectively, this tube is the real deal. The speakers in my ham shack are some little Yamaha monitors. The room is 8 x 12. Where the 801A begins to break up on bass, the SV 811 hunkers down and shakes the walls. Literally. This is a 65 watt tube, and has at least another 2 to 3 dB left in it, and you can really hear it.
I see no reason why 15 to 20 watts SE cannot be realized from this tube, on this board.
I see several ways to get there. SV-811-3 would be more compatible with the unmodified bias circuit and would almost certainly get there, but might present drive problems with the low mu. I don't have any, and they are expensive, so I am unlikely to acquire any.
The bias circuit could be modified to present a lower bias voltage, but this might make the circuit incompatible with the 801A. This might be viable.
The B+ could be voltage doubled and then adjusted, but again this might make the circuit incompatible with 801A which has a 600 volt limit, IIRC. This also might be viable.
An external custom power supply may be the best option and would also be useful for crushing 6146's.
With so many projects in the pipeline, I may just enjoy this modest amount of success before I come back to this tube on this board.
Happy New Year to all.
Win W5JAG
Attachments
I have one of those. Mine is marked RCA 5AR4. I have used it in several amps for about 20 years. So far it's indestructible, even when I screwed up and put a 100uF in for the first filter cap.
Good grief, one step forward and two steps back.
Powered it up last night to listen to a ball game, and the audio was being strongly modulated by line ac. The ac has an odd, warbly, out of phase, quality, and goes up and down with the volume, as if it is coming in on the grid, maybe both the grid and the plate.
Which is certainly possible, as the grid supply uses taps off of the main HV winding.
The first thing that came to my mind, and the way my luck has run with this, was that the PT has a shorted turn, or turns, on the HV secondary. But here's the weird part: I have excessive B+ voltage, by a lot - it's hard to keep it below 500 vdc, when it's fully loaded up. Normal unloaded ac voltage looks ok - 370 - 0 - 370, within a volt or two each side of CT.
The new filament circuit looks to be working nominally, I suspected that first since I had just fooled with it, but the voltages are unchanged, and all the diodes get hot, so they are conducting. Not sure how a malfunction here could cause a 10% increase in loaded B+.
I isolated a couple of windings and did some resistance checks on the PT - they look ok. No windings are shorted to another, and resistances look nominal.
Choke is speced at 80 ohms, and I measure 76. I don't see much voltage drop across it - only a few volts, IIRC, so I can't say this is ruled out.
I've ruled out tubes.
I have not checked the electrolytics on the HV supply - not sure how a failure here could boost the B+. So, I'm thinking, shorted turn, but not sure how that would boost the B+.
Anyone have any other ideas?
Win W5JAG
Powered it up last night to listen to a ball game, and the audio was being strongly modulated by line ac. The ac has an odd, warbly, out of phase, quality, and goes up and down with the volume, as if it is coming in on the grid, maybe both the grid and the plate.
Which is certainly possible, as the grid supply uses taps off of the main HV winding.
The first thing that came to my mind, and the way my luck has run with this, was that the PT has a shorted turn, or turns, on the HV secondary. But here's the weird part: I have excessive B+ voltage, by a lot - it's hard to keep it below 500 vdc, when it's fully loaded up. Normal unloaded ac voltage looks ok - 370 - 0 - 370, within a volt or two each side of CT.
The new filament circuit looks to be working nominally, I suspected that first since I had just fooled with it, but the voltages are unchanged, and all the diodes get hot, so they are conducting. Not sure how a malfunction here could cause a 10% increase in loaded B+.
I isolated a couple of windings and did some resistance checks on the PT - they look ok. No windings are shorted to another, and resistances look nominal.
Choke is speced at 80 ohms, and I measure 76. I don't see much voltage drop across it - only a few volts, IIRC, so I can't say this is ruled out.
I've ruled out tubes.
I have not checked the electrolytics on the HV supply - not sure how a failure here could boost the B+. So, I'm thinking, shorted turn, but not sure how that would boost the B+.
Anyone have any other ideas?
Win W5JAG
Been out of town for a few days, so this afternoon is the first opportunity to troubleshoot this thing.
To eliminate the PT, I just disconnected the existing PT, and jumpered in an inadequate sub. This fixed the overvoltage, where I can work on it. Still sounds like crap.
Got the scope out. Here's what I think is good: everything up to and including the grids of the 801's. Put the scope on the output, and a sine wave distorts pretty bad - looks similar to a triangle wave on the positive side of the wave. A 1kHz square wave looks identical to the third pic above.
Win W5JAG
To eliminate the PT, I just disconnected the existing PT, and jumpered in an inadequate sub. This fixed the overvoltage, where I can work on it. Still sounds like crap.
Got the scope out. Here's what I think is good: everything up to and including the grids of the 801's. Put the scope on the output, and a sine wave distorts pretty bad - looks similar to a triangle wave on the positive side of the wave. A 1kHz square wave looks identical to the third pic above.
Win W5JAG
I also have some marked with RCA trade dress, in RCA boxes, but marked as 5V4GA on the box and tube. This particular one has the regular consumer Sylvania trade dress, and came out of a consumer Sylvania box marked 5V4GA.
I have some Sylvania 5AR4's that are obvious Mullards. I wonder if at some point in time, the Mullard's went out of production, and Sylvania came up with this design to replace them, and just dropped production of the real 5V4GA. Thus, when the Feds needed 5V4GA's, this 5AR4 got the JAN designation, because it was what was left in production.
However, I have never seen one of these tubes in a box actually marked as a 5AR4 on the box. IIRC, all the JAN boxes I have seen were marked as 5V4GA. They may exist, but I haven't seen one.
But they are clearly 5AR4's.
It took me a while to troubleshoot the problem to the filter caps ( well, I also watched some football ) because I was having a hard time accepting they were bad, and I don't understand the relationship of the excessive B+ with their failure. The ac ripple, I get, but not the high B+.
These were new caps, rated 500 volts, and by the standards of this sub forum, comparatively unabused. They looked and smelled OK. I also replaced a bad one with a bad one, but it let off that characteristic stink, although it did not explode.
I hope the drama is over for a while, knock on wood.
Win W5JAG
The Tusotek meets the ( modifed ) TSE
They are sold on eBay under the Tusotek brand, and are very inexpensive. There seems to be two types - one a voltage regulator only device, and the other a constant current regulator and voltage regulator device. The price difference is often only a few cents. This note concerns the constant current / constant voltage device.
I set it up on the bench last week, and looked at it on the scope. When powering two 801A's, it did show some switching noise at 120 KHz, but seemed otherwise satisfactory. Running it from a 13.8 volt homebew linear supply, it seemed quite cool and held voltage to 0.01 volt over the course of a day, as measured by a cheap HF VOM. It also had no issues powering a single 572B for several hours. Generated heat was insignificant with the load of the 572B.
I hastily installed it in the modified TSE - with no attempt to shield anything from it, or vice versa. I expected squeals, whistles, hiss, hum or some other unsatisfactory operation. To my surprise, the amp was stone cold silent. No audible issues. Additionally, I did not have to touch the adjustments that had been set up on the bench with the 13.8 volt source. It accepted the output from the rectifier and delivered 7.5 volts at the proper current. It would have been in buck mode on the bench, but is probably on the threshold of buck / boost as installed in the modified TSE. It didn't care. Further, because of the way I have the TSE modified, and this quickie installation, input to the Tusotek is raw unfiltered DC. It doesn't care. There is a 10,000 uF cap on the output. It doesn't seem to care about that, either. I haven't looked at it on the scope as installed - it seems fine as - is.
The implications are obvious. As installed, this one module can likely power a pair of 45, 46, and 47 as well as the 801's, with nothing more than a simple screwdriver adjustment for the voltage and current. It may do a pair of 2A3's, although that would be at the current limit for a single module. Two could power a pair of SV-811 or SV-572.
As installed, heat dissipation may be an issue when it was not on the bench. So far, it has not been. Early days, but it looks like promising.
Win W5JAG
They are sold on eBay under the Tusotek brand, and are very inexpensive. There seems to be two types - one a voltage regulator only device, and the other a constant current regulator and voltage regulator device. The price difference is often only a few cents. This note concerns the constant current / constant voltage device.
I set it up on the bench last week, and looked at it on the scope. When powering two 801A's, it did show some switching noise at 120 KHz, but seemed otherwise satisfactory. Running it from a 13.8 volt homebew linear supply, it seemed quite cool and held voltage to 0.01 volt over the course of a day, as measured by a cheap HF VOM. It also had no issues powering a single 572B for several hours. Generated heat was insignificant with the load of the 572B.
I hastily installed it in the modified TSE - with no attempt to shield anything from it, or vice versa. I expected squeals, whistles, hiss, hum or some other unsatisfactory operation. To my surprise, the amp was stone cold silent. No audible issues. Additionally, I did not have to touch the adjustments that had been set up on the bench with the 13.8 volt source. It accepted the output from the rectifier and delivered 7.5 volts at the proper current. It would have been in buck mode on the bench, but is probably on the threshold of buck / boost as installed in the modified TSE. It didn't care. Further, because of the way I have the TSE modified, and this quickie installation, input to the Tusotek is raw unfiltered DC. It doesn't care. There is a 10,000 uF cap on the output. It doesn't seem to care about that, either. I haven't looked at it on the scope as installed - it seems fine as - is.
The implications are obvious. As installed, this one module can likely power a pair of 45, 46, and 47 as well as the 801's, with nothing more than a simple screwdriver adjustment for the voltage and current. It may do a pair of 2A3's, although that would be at the current limit for a single module. Two could power a pair of SV-811 or SV-572.
As installed, heat dissipation may be an issue when it was not on the bench. So far, it has not been. Early days, but it looks like promising.
Win W5JAG
Attachments
Last edited:
I'm pretty impressed with these inexpensive buck / boost modules - enough so to order four more.
One of the Tusotek modules can power a pair of 2A3. The specified current limit is 5 amps, and adjustment of the current set is a bit tricky with a pair of 2A3. There is no documentation with these, and although advertised as a current regulator, it may be a current limiter - I'm not sure.
Start up is a bit dicey with the 5 amp load - the module hits about half voltage right away, then as the cathodes warm and pull current, a few seconds later it hits the specified 2.5 volts. I haven't tried to measure the actual start up current draw of a pair of 2A3 - when cold, a pair might be more than five amps. The module may be limiting at 5 amps, or regulating, I just don't know.
Regardless, once it starts, it works fine. The amp is stone cold silent on my small Yamaha's - haven't had a chance to try it on the 96dB Klipsch, but I do not expect any issues. No audible artifacts that I can hear. I've run it as long as 24 hours continuous with a pair of 2A3 off a 6.3 volt filament circuit. They are so efficient, heat just does not seem to be a problem.
So far I have used 45, 2A3, and 801A with the module - a spread from 7.5 volts at 2.5 amps to 2.5 volts at 5 amps, with only a couple of screwdriver adjustments. With the tapped power transformer, multiple impedance Hammond OPT's, and the Tusotek, this may be an accidental universal TSE.
Win W5JAG
One of the Tusotek modules can power a pair of 2A3. The specified current limit is 5 amps, and adjustment of the current set is a bit tricky with a pair of 2A3. There is no documentation with these, and although advertised as a current regulator, it may be a current limiter - I'm not sure.
Start up is a bit dicey with the 5 amp load - the module hits about half voltage right away, then as the cathodes warm and pull current, a few seconds later it hits the specified 2.5 volts. I haven't tried to measure the actual start up current draw of a pair of 2A3 - when cold, a pair might be more than five amps. The module may be limiting at 5 amps, or regulating, I just don't know.
Regardless, once it starts, it works fine. The amp is stone cold silent on my small Yamaha's - haven't had a chance to try it on the 96dB Klipsch, but I do not expect any issues. No audible artifacts that I can hear. I've run it as long as 24 hours continuous with a pair of 2A3 off a 6.3 volt filament circuit. They are so efficient, heat just does not seem to be a problem.
So far I have used 45, 2A3, and 801A with the module - a spread from 7.5 volts at 2.5 amps to 2.5 volts at 5 amps, with only a couple of screwdriver adjustments. With the tapped power transformer, multiple impedance Hammond OPT's, and the Tusotek, this may be an accidental universal TSE.
Win W5JAG
Learned enough about using the tusotek, 801A, and SV-811 to go ahead and tear the amp down for a much needed overhaul / rebuild.
Not yet sure what it will come back as. The DHT modified SSE can do a good enough job with small stuff like 2A3, 45, 46, 47, etc., so I will probably stay with thoriated tungsten for this one.
Win W5JAG
Not yet sure what it will come back as. The DHT modified SSE can do a good enough job with small stuff like 2A3, 45, 46, 47, etc., so I will probably stay with thoriated tungsten for this one.
Win W5JAG
- Status
- Not open for further replies.