My "shop" Simple SE amp (with the outboard sockets for 6AV5 incandescence experiments) wants KT-88s so to play that Rock & Roll Music. I sense some jealousy of the ratty old QUAD 303.
For this mini-project, I would like to get the B+ up around 470-475vdc using current iron:
PSUD-II says that the options for a rectifier are
(1) Solid State. Nah, yeah, nah...
(2) GZ-37. HA! you make me laugh, PSUD-II....
(3) 2 x 5AR4 (GZ34). Sorry, bubba, not enuf fils current....
(4) "Graetz" 5AR4, which I believe means HYBRID rectification with 4007s on the HV terminals. IF SO, that would work great..
I assume that the 5AR4GRAETZ is like the famous DYNACO "yellow sheet" mod shown below, IIRC necessitated by the dodginess of current-production 5AR4s (have flashed TWO of them myself, one audio, one guitar, and swore them off about 5 years ago).
Is this indeed what 5AR4GRAETZ means in the PSUD-II rectifier list? Can't tell from the limited specs: PIV 1500V, IFSM 99A, IFRM 1.5A. Hopefully it doesn't mean a FULL BRIDGE S/S rectifier in front of the tube...
If so, has anybody used this mod with 5AR4s on TubeLab boards?
I assume it could be done by simply putting UF-4007s in the terminal blocks and soldering the HV leads to the diodes (with nice red shrink wrap, of course). Is that right?
In other news, I played around with a "lower impact" P/S topology using TWO surprisingly tiny chokes I have lying around (Hammond 193 H, 5H, 200mA, 65 ohms DCR). See schematic below. This gives about 470 vdc B+, which is in the ballpark.
The idea would be to use a terminal strip between the chokes, and bring the 47uF cap back to the Star Ground.
Generally, I really like 3 stage filters with low-medium-high capacitance; in this case, the PSU-II shows virtually NO ripple (eyeballing about 0.0something volts RMS; my eyeball has a square root function, but with loose precision).
If a second choke isn't available or won't fit, an RC filter works pretty well, too. I may change out the chokes simply to get the 200mA current rating.
If you're using a choke on your board ANYWAY, it's a simple mod to add a third stage to your power supply filter, and then "tone-down" each stage...
For this mini-project, I would like to get the B+ up around 470-475vdc using current iron:
- PST is prewar Thordarson 390-0-390v, 200mA.
- Choke is Thordarson "swinging" choke, measures 8.5H at 100 ohms DCR
PSUD-II says that the options for a rectifier are
(1) Solid State. Nah, yeah, nah...
(2) GZ-37. HA! you make me laugh, PSUD-II....
(3) 2 x 5AR4 (GZ34). Sorry, bubba, not enuf fils current....
(4) "Graetz" 5AR4, which I believe means HYBRID rectification with 4007s on the HV terminals. IF SO, that would work great..
I assume that the 5AR4GRAETZ is like the famous DYNACO "yellow sheet" mod shown below, IIRC necessitated by the dodginess of current-production 5AR4s (have flashed TWO of them myself, one audio, one guitar, and swore them off about 5 years ago).
Is this indeed what 5AR4GRAETZ means in the PSUD-II rectifier list? Can't tell from the limited specs: PIV 1500V, IFSM 99A, IFRM 1.5A. Hopefully it doesn't mean a FULL BRIDGE S/S rectifier in front of the tube...
If so, has anybody used this mod with 5AR4s on TubeLab boards?
I assume it could be done by simply putting UF-4007s in the terminal blocks and soldering the HV leads to the diodes (with nice red shrink wrap, of course). Is that right?
In other news, I played around with a "lower impact" P/S topology using TWO surprisingly tiny chokes I have lying around (Hammond 193 H, 5H, 200mA, 65 ohms DCR). See schematic below. This gives about 470 vdc B+, which is in the ballpark.
The idea would be to use a terminal strip between the chokes, and bring the 47uF cap back to the Star Ground.
Generally, I really like 3 stage filters with low-medium-high capacitance; in this case, the PSU-II shows virtually NO ripple (eyeballing about 0.0something volts RMS; my eyeball has a square root function, but with loose precision).
If a second choke isn't available or won't fit, an RC filter works pretty well, too. I may change out the chokes simply to get the 200mA current rating.
If you're using a choke on your board ANYWAY, it's a simple mod to add a third stage to your power supply filter, and then "tone-down" each stage...
Attachments
Last edited:
Addendum: some tedious searching lead to this post on Hybrid Rectification on the TSE-II
TSEII with a Hybrid brigde?
I had forgotten about the Bias Supply on the TSE/TSE-II boards; diodes would certainly screw that up.
In any case, to be clear, I am talking about the TubeLab Simple SE (SSE) board, which is cathode biased.
TSEII with a Hybrid brigde?
I had forgotten about the Bias Supply on the TSE/TSE-II boards; diodes would certainly screw that up.
In any case, to be clear, I am talking about the TubeLab Simple SE (SSE) board, which is cathode biased.
Last edited:
The SSE has provisioning for those diodes as well as an NTC.
I think the board was revised at some stage and these components are now part of it, at least mine has them.
I think the board was revised at some stage and these components are now part of it, at least mine has them.
SSE Diodes
Do you mean THESE FREDs?
The "Graetz" mod (apparently popularized by a guy named Eli Duttman) comprises one diode on each red HV input from the PST, "pointed-in". To avoid confusion, I call it the Duttman/DYNACO mod.
It's like two half-wave rectifiers in series; the SS diodes "do the heavy lifting", while the 5AR4 looks pretty and supposedly smoothes-out the hash from the 4007s.
The FREDs on the SSE board are an alternative rectifier (ie you can have a tube rectifier -OR- a Solid State rectifier, but not both). Can't see any way to use those circuit board positions shy of cutting and re-routing traces.
Note that the nomenclature seems to be a little confusing, since the original Graetz design was a solid-state full bridge rectifier feeding a vacuum tube rectifier, if I am understanding correctly. This is why I am asking the question: what does the 5AR4GRAETZ in PSUD-II represent, a SS bridge + tube or a SS Half-Wave + tube?
Do you mean THESE FREDs?
The "Graetz" mod (apparently popularized by a guy named Eli Duttman) comprises one diode on each red HV input from the PST, "pointed-in". To avoid confusion, I call it the Duttman/DYNACO mod.
It's like two half-wave rectifiers in series; the SS diodes "do the heavy lifting", while the 5AR4 looks pretty and supposedly smoothes-out the hash from the 4007s.
The FREDs on the SSE board are an alternative rectifier (ie you can have a tube rectifier -OR- a Solid State rectifier, but not both). Can't see any way to use those circuit board positions shy of cutting and re-routing traces.
Note that the nomenclature seems to be a little confusing, since the original Graetz design was a solid-state full bridge rectifier feeding a vacuum tube rectifier, if I am understanding correctly. This is why I am asking the question: what does the 5AR4GRAETZ in PSUD-II represent, a SS bridge + tube or a SS Half-Wave + tube?
Attachments
Which Gaetz Hybrid?
To add ANOTHER level of complexity, perhaps the 5AR4GRAETZ refers to a "grounded diode" version. Sort of looks like an artificial center-tap to me.
See: SETA - Single Ended Tube Amplifier
Again, that would be easy to do; the HV input is very close to the Star Ground.
I suppose that it would mean that the Transformer Center Tap would have to be lifted off ground.
Again, any advice about what the 5AR4GRAETZ model represents in PSUD-II?
(a) SS Bridge into vacuum tube rectifier (I doubt this one)
(b) SS diodes in series with the vacuum tube rectifier (with unchanged HV center tap to ground); the "DYNACO/Duttman" mod.
(c) SS diodes between vacuum tube rectifier plates and GROUND, with PST HV center tap LIFTED off ground...
The 5AR4Graetz simulation looks great, just don't know how to build it.
I continue the search.
To add ANOTHER level of complexity, perhaps the 5AR4GRAETZ refers to a "grounded diode" version. Sort of looks like an artificial center-tap to me.
See: SETA - Single Ended Tube Amplifier
Again, that would be easy to do; the HV input is very close to the Star Ground.
I suppose that it would mean that the Transformer Center Tap would have to be lifted off ground.
Again, any advice about what the 5AR4GRAETZ model represents in PSUD-II?
(a) SS Bridge into vacuum tube rectifier (I doubt this one)
(b) SS diodes in series with the vacuum tube rectifier (with unchanged HV center tap to ground); the "DYNACO/Duttman" mod.
(c) SS diodes between vacuum tube rectifier plates and GROUND, with PST HV center tap LIFTED off ground...
The 5AR4Graetz simulation looks great, just don't know how to build it.
I continue the search.
Attachments
Even more confusion : in Europe "Graetz Circuit" is a name for the standard four diode rectifier.
Further thoughts: If the PSUD-II "5AR4Graetz" describes a "hybrid bridge" then the Grounded Diode (version "C") must be what they mean. Looks exactly like a bridge rectifier, except with 2 vacuum-tube diodes on the "output".
This would of course require lifting the SSE Center Tap.
HOWEVER, it would also permit using a non-center tapped PST with a vacuum tube rectifier. I seem to remember a couple of those non-CT PSTs lurking in the back of the transformer shelf...
This would of course require lifting the SSE Center Tap.
HOWEVER, it would also permit using a non-center tapped PST with a vacuum tube rectifier. I seem to remember a couple of those non-CT PSTs lurking in the back of the transformer shelf...
Each plate of the GZ34 has a diode in front of it like the "Yellow Dynaco Mod". The Centre tap of the H.V. has an NTC. Its all on the board now. It just takes the initial surge off the GZ34 at turn on, hopefully preventing the flashover and resulting death of the GZ34 which I have experienced with a J.J. in another amp (TSEII)
I bought 2 SSE boards. One for a keeper and one to play with.
Working on the player now. 6883A powered, fixed bias and regulated G2. Now trying to repurpose a chassis yet again. Hopefully will finish within a month or two 😉
I bought 2 SSE boards. One for a keeper and one to play with.
Working on the player now. 6883A powered, fixed bias and regulated G2. Now trying to repurpose a chassis yet again. Hopefully will finish within a month or two 😉
There is some conflicting information scattered through this thread. I'm hoping that I can make things clearer, not more confusing.
Andrewbee states that the "yellow sheet" mod is built into the SSE board. That's only true for boards made after mid 2010. The older boards did not have the diodes or the CL140. They were added due to some rather poor quality from all three vendors of new production 5AR4's at the time. Some builders reported one or even two brand new 5AR4's would spark out on first power up, but once a good one was found it would live for years. These parts can be added in series with the transformer wires. The diodes improve the PIV of the tube, and the CL-140 gives the cathode a few more seconds to heat before it tries to fill the empty filter cap when it's only half hot. The thin spots in the cathode coating heat the fastest and try to pass all the current, often unsuccessfully.
The Graetz bridge or Graetz rectifier usually refers to a typical 4 diode bridge, built with silicon or tubes. A hybrid Graetz refers to one that's half silicon, half tube. The yellow sheet mod is NOT a bridge, it's just two diodes in series on each leg of a conventional FWCT rectifier.
Using a bridge on a 390-0-390 transformer will result in about 950 volts of B+, far too much for a typical SSE build.
It is possible to run an SSE from a non CT power transformer. Connect the transformer secondary to the red terminals, and wire a diode from each red terminal to ground. If there is a CT on the transformer, it must not be connected to anything. B+ will be about 1.3 X the secondary voltage.
Andrewbee states that the "yellow sheet" mod is built into the SSE board. That's only true for boards made after mid 2010. The older boards did not have the diodes or the CL140. They were added due to some rather poor quality from all three vendors of new production 5AR4's at the time. Some builders reported one or even two brand new 5AR4's would spark out on first power up, but once a good one was found it would live for years. These parts can be added in series with the transformer wires. The diodes improve the PIV of the tube, and the CL-140 gives the cathode a few more seconds to heat before it tries to fill the empty filter cap when it's only half hot. The thin spots in the cathode coating heat the fastest and try to pass all the current, often unsuccessfully.
The Graetz bridge or Graetz rectifier usually refers to a typical 4 diode bridge, built with silicon or tubes. A hybrid Graetz refers to one that's half silicon, half tube. The yellow sheet mod is NOT a bridge, it's just two diodes in series on each leg of a conventional FWCT rectifier.
Using a bridge on a 390-0-390 transformer will result in about 950 volts of B+, far too much for a typical SSE build.
It is possible to run an SSE from a non CT power transformer. Connect the transformer secondary to the red terminals, and wire a diode from each red terminal to ground. If there is a CT on the transformer, it must not be connected to anything. B+ will be about 1.3 X the secondary voltage.
Great stuff George, thanks.
My SSE boards are pre-2010, and I ate at least one JJ 5AR4 on start-up.
Ate another one in a guitar amp. Apparently I wasn't the only one...
Also didn't know about the CL140; will research and add that to my amps (I use CLs commonly nowadays, but didn't when I built the SSEs).
Last question (well, tonight anyway): there's a rectifier entry in PSUD-II called "5AR4GRAETZ". I can't tell from the underlying text file WHAT configuration this refers to. Can anybody tell me if this a Hybrid Bridge, ie a 5AR4 and 2 x 4007 diodes?
Whatever it is, it models great for KT-88 at about 470v B+
My SSE boards are pre-2010, and I ate at least one JJ 5AR4 on start-up.
Ate another one in a guitar amp. Apparently I wasn't the only one...
Also didn't know about the CL140; will research and add that to my amps (I use CLs commonly nowadays, but didn't when I built the SSEs).
Last question (well, tonight anyway): there's a rectifier entry in PSUD-II called "5AR4GRAETZ". I can't tell from the underlying text file WHAT configuration this refers to. Can anybody tell me if this a Hybrid Bridge, ie a 5AR4 and 2 x 4007 diodes?
Whatever it is, it models great for KT-88 at about 470v B+
My original "industrial" looking SSE used the old board without diodes or Inrush Current Limiters. The old Sylvania made, RCA branded 5AR4 that was in it still lives in a TSE-II over 15 years later.
In late 2009 people started complaining about blown rectifier tubes, mostly Sovtek's. I had been using Sovtek tubes without issue in other amps, but mine were 3+ years old, so I got some new ones, and sure enough one popped on first power up. I added a CL90 in series with the power transformer primary, a CL140 in the HV secondary CT, and the two 1N4007 diodes, and no more rectifier problems occurred. Later in 2010 there were bad batches of JJ and Shuguang tubes in circulation.
I know that GE used to sample test about 50 tubes out of a production lot. If the failure rate was above a certain threshold, the entire lot was "scrapped." RCA had a similar procedure, but it was not made public. I suspect that other old manufacturers did something similar, since it would be impossible to test every tube that left the factory due to the heater warm up time. I suspect that the modern new production vendors do the same.
Many US, and other "premium tube" vendors buy tubes in bulk from these manufacturers, burn them in, test, and maybe match them, and sell them for a premium price. The "bad" ones get "scrapped."
Where do the scrap tubes go?
In the old days they were bought by the second tier tube sellers, who may or may not actually test them, then put their own brand on them and sell them.
In the late 60's a new GE 6L6GC cost about $4 at a local electronics store. I could go to the Eagle Army - Navy store and get a "Thoro Test" 6L6GC for 99 cents. It was one of those rejected GE tubes. I had a pair in one of my DIY guitar amps which I beat on hard nearly every day. They lived for several years.
About 15 years ago Shuguang made "Coke Bottle" 6L6GC's. I bought a pair from AES since they claimed that "we couldn't blow up this tube no matter how hard we tried." I think they were about $8 each at the time. Guess, what, my pair are still alive, despite trying pretty hard!
I found some of the same tubes on Ebay for $3 each, and they sucked, mostly due to poor alignment causing hot spots on the plate. They had to be reject tubes at the factory.
I'm guessing that the reject 5AR4's from all three vendors mad their way into normal distribution during 2010. After blasting a couple Chinese 5AR4 tubes in my 845 SE amp, I bought 4 Sovteks from a tube vendor who tested them and "guaranteed" them. They worked fine, and are still in the amp.
In late 2009 people started complaining about blown rectifier tubes, mostly Sovtek's. I had been using Sovtek tubes without issue in other amps, but mine were 3+ years old, so I got some new ones, and sure enough one popped on first power up. I added a CL90 in series with the power transformer primary, a CL140 in the HV secondary CT, and the two 1N4007 diodes, and no more rectifier problems occurred. Later in 2010 there were bad batches of JJ and Shuguang tubes in circulation.
I know that GE used to sample test about 50 tubes out of a production lot. If the failure rate was above a certain threshold, the entire lot was "scrapped." RCA had a similar procedure, but it was not made public. I suspect that other old manufacturers did something similar, since it would be impossible to test every tube that left the factory due to the heater warm up time. I suspect that the modern new production vendors do the same.
Many US, and other "premium tube" vendors buy tubes in bulk from these manufacturers, burn them in, test, and maybe match them, and sell them for a premium price. The "bad" ones get "scrapped."
Where do the scrap tubes go?
In the old days they were bought by the second tier tube sellers, who may or may not actually test them, then put their own brand on them and sell them.
In the late 60's a new GE 6L6GC cost about $4 at a local electronics store. I could go to the Eagle Army - Navy store and get a "Thoro Test" 6L6GC for 99 cents. It was one of those rejected GE tubes. I had a pair in one of my DIY guitar amps which I beat on hard nearly every day. They lived for several years.
About 15 years ago Shuguang made "Coke Bottle" 6L6GC's. I bought a pair from AES since they claimed that "we couldn't blow up this tube no matter how hard we tried." I think they were about $8 each at the time. Guess, what, my pair are still alive, despite trying pretty hard!
I found some of the same tubes on Ebay for $3 each, and they sucked, mostly due to poor alignment causing hot spots on the plate. They had to be reject tubes at the factory.
I'm guessing that the reject 5AR4's from all three vendors mad their way into normal distribution during 2010. After blasting a couple Chinese 5AR4 tubes in my 845 SE amp, I bought 4 Sovteks from a tube vendor who tested them and "guaranteed" them. They worked fine, and are still in the amp.
Attachments
Still VERY Confused about "5AR4Graetz"
This is good stuff, George, but I'm still confused about the "5AR4Graetz" half-wave rectifier in PSUD-II.
Here are the B+ voltages from the Duncan PSUD-II, with 390-0-390v secondaries, my proposed filter, and 180mA Load
(ie apples-to-apples comparison):
5AR4: ~444V mean (but "IFRM of 0.75A exceeded at 0.77A")
"5AR4Graetz": ~458V mean (no error messages)
FullWave 1N4007: ~464V mean ("PIV of 1000 exceeded at 1004"
The "5AR4Graetz" (under Half-Wave tube rectifiers) is acting like a half-wave rectifier, NOT like a bridge, both in B+ voltage and in the shape of the voltage curve.
It IS acting like a "hybrid" (but a Full-Wave hybrid) in that it tolerates higher IFRM than the 5AR4 alone, and doesn't see the high PIV of the Full Wave 1N4007s alone.
Rectifier Specs comparison:
1N4007 Full Wave: 1000V PIV, 30A IFSM, 6A IFRM
5AR4Graetz: 1500V PIV, 99A IFSM, 1.5A IFRM
5AR4: 1500V PIV, 99A IFSM, 0.75A IFRM
I think "5AR4Graetz" has to be the Yellow-Sheet Dynaco Mod (4007s in series), and NOT a Hybrid Bridge.
This is good stuff, George, but I'm still confused about the "5AR4Graetz" half-wave rectifier in PSUD-II.
Here are the B+ voltages from the Duncan PSUD-II, with 390-0-390v secondaries, my proposed filter, and 180mA Load
(ie apples-to-apples comparison):
5AR4: ~444V mean (but "IFRM of 0.75A exceeded at 0.77A")
"5AR4Graetz": ~458V mean (no error messages)
FullWave 1N4007: ~464V mean ("PIV of 1000 exceeded at 1004"
The "5AR4Graetz" (under Half-Wave tube rectifiers) is acting like a half-wave rectifier, NOT like a bridge, both in B+ voltage and in the shape of the voltage curve.
It IS acting like a "hybrid" (but a Full-Wave hybrid) in that it tolerates higher IFRM than the 5AR4 alone, and doesn't see the high PIV of the Full Wave 1N4007s alone.
Rectifier Specs comparison:
1N4007 Full Wave: 1000V PIV, 30A IFSM, 6A IFRM
5AR4Graetz: 1500V PIV, 99A IFSM, 1.5A IFRM
5AR4: 1500V PIV, 99A IFSM, 0.75A IFRM
I think "5AR4Graetz" has to be the Yellow-Sheet Dynaco Mod (4007s in series), and NOT a Hybrid Bridge.