LOL
I have to say that's tempting
I have to say that's tempting
Dynaco ST-35 Build – DG Style
- Thread starter njcanuck
- Start date
If you don't want to double up your resistors:
http://www.mouser.com/Passive-Components/Resistors/_/N-5g9n?P=1yzeg4tZ1yze5qqZ1z0z819
-D
http://www.mouser.com/Passive-Components/Resistors/_/N-5g9n?P=1yzeg4tZ1yze5qqZ1z0z819
-D
When you order from mouser in small amounts you can request that they send it USPS first class. I had to just do that for a couple of resistors and the shipping was like $2.50. It is not necessarily apparent on their order forms but you can add additional shipping detail and ask for the cheaper shipping option.
njcanuck
aka: MacGyver
That is good to know. Thanks
njcanuck
aka: MacGyver
Ok, got this puppy wired up and singing! 
The wiring went reasonably smoothly with only a couple of minor glitches. I worked mostly from the diagrams as neither the Dynakit instructions nor DG’s instructions were intended for a scratch build clone with EFB. As a result I missed a couple of connections on the PC-13 boards.
One is still confusing me: On the PC-13 board there is clearly shown a connection on the lower 6BQ5 from Pin 3 to Pin 4 (circled on Pic 1 below). I made this connection and it works well. However, I can’t see this connection on the schematic (Pic 2 below)? It looks to me like this connection bypasses the 10 ohm resistors. What am I not seeing or missing here???
Another small “gotcha” is that the input jacks need to be grounded to the chassis (as in the original design). If fancy insulated jacks, like mine, are used without a proper ground, the bias runs wild and smokes the power supply.
Thank god for variacs, ammeters and quick-kill switches! The problem, solution and explanation was found in this thread: http://www.audiokarma.org/forums/index.php?threads/st35-with-efb-bias-problem.697612/ In Pic 3 you can see the black ground wires I added to the jacks.
The wiring went reasonably smoothly with only a couple of minor glitches. I worked mostly from the diagrams as neither the Dynakit instructions nor DG’s instructions were intended for a scratch build clone with EFB. As a result I missed a couple of connections on the PC-13 boards.
One is still confusing me: On the PC-13 board there is clearly shown a connection on the lower 6BQ5 from Pin 3 to Pin 4 (circled on Pic 1 below). I made this connection and it works well. However, I can’t see this connection on the schematic (Pic 2 below)? It looks to me like this connection bypasses the 10 ohm resistors. What am I not seeing or missing here???
Another small “gotcha” is that the input jacks need to be grounded to the chassis (as in the original design). If fancy insulated jacks, like mine, are used without a proper ground, the bias runs wild and smokes the power supply.
Thank god for variacs, ammeters and quick-kill switches! The problem, solution and explanation was found in this thread: http://www.audiokarma.org/forums/index.php?threads/st35-with-efb-bias-problem.697612/ In Pic 3 you can see the black ground wires I added to the jacks.
njcanuck
aka: MacGyver
Here’s a couple of topside views, the second one showing it hooked up to dummy loads, voltmeters, sig gen, scope, etc.
Matching tubes in this unit is a breeze. Dave pointed out to me that EFB will let you run the amp with just one single 6BQ5 installed. So simply warm up the amp with just one installed and record the bias voltage. Do this for each output tube in turn without changing any settings and using the same socket. Those with the closest measurements are the closest matched. Pick your best pairs for each channel (within reason) and you’re good to go.
At the moment, the feedback network for 16 ohm taps is still installed but even so, the square waves are pretty impressive. The scope pics are showing a quick and dirty look at 2KHz and 10KHz respectively. Not bad at all! Sometime next week I’ll measure frequency response, power output, etc. By then I should have the parts for the 8 ohm networks.
For now I’m just listening. Had it running for a couple of hours this afternoon and it was pretty sweet!
Edit: Oops, I must have screwed up that 10KHz measurement somehow. See post #85 for the correct and not-so-nice plot.
Matching tubes in this unit is a breeze. Dave pointed out to me that EFB will let you run the amp with just one single 6BQ5 installed. So simply warm up the amp with just one installed and record the bias voltage. Do this for each output tube in turn without changing any settings and using the same socket. Those with the closest measurements are the closest matched. Pick your best pairs for each channel (within reason) and you’re good to go.
At the moment, the feedback network for 16 ohm taps is still installed but even so, the square waves are pretty impressive. The scope pics are showing a quick and dirty look at 2KHz and 10KHz respectively. Not bad at all! Sometime next week I’ll measure frequency response, power output, etc. By then I should have the parts for the 8 ohm networks.
For now I’m just listening. Had it running for a couple of hours this afternoon and it was pretty sweet!
Edit: Oops, I must have screwed up that 10KHz measurement somehow. See post #85 for the correct and not-so-nice plot.
Last edited:
Those are fantastic looking 10kc waves.
3 is the cathode, 4 is the heater. Tying them together puts some DC on the heater circuit. Usually you'd do a 100-470 ohm resistor from each heater leg to the cathode to provide a virtual center tap to ground for the heater circuit, or tie it to the actual center tap of the transformer if its got one. It'll work unbalanced though, or maybe there is another reason for it being done that way.
3 is the cathode, 4 is the heater. Tying them together puts some DC on the heater circuit. Usually you'd do a 100-470 ohm resistor from each heater leg to the cathode to provide a virtual center tap to ground for the heater circuit, or tie it to the actual center tap of the transformer if its got one. It'll work unbalanced though, or maybe there is another reason for it being done that way.
Nice job Rich! As Gadget pointed out, the original connection between pins 3&4 at one output tube on each board was to bias the heater circuit above ground by the amount of the output tube bias voltage. If this were a high sensitivity amplifier (i.e., eith phono preamps, etc.) then applying it in a balanced format would be important. But in a low sensitivity setting such as a power amp, it really isn't required. Audibly, you likely would not be able to tell the difference between a balanced and unbalanced heater bias connection.
Now with EFB installed, the jumper at each tube should theoretically be made between pin 4 of one output tube, and the output of the EFB regulator circuit (where the black Com test terminal is connected to). However, because it is in fact an unbalanced heater circuit, one side of the heater circuit can still be connected straight to a cathode (to obtain the heater bias voltage), and there will be no modulation of the cathode voltage by the AC heater supply, because the other side of the heater supply does not reference ground in any way (which would short out the EFB bias if it did!). As well, the connection won't unset the output bias indication, either. So for all practical purposes, the original jumper connection is just fine.
Congrats on getting it going!
Dave
Now with EFB installed, the jumper at each tube should theoretically be made between pin 4 of one output tube, and the output of the EFB regulator circuit (where the black Com test terminal is connected to). However, because it is in fact an unbalanced heater circuit, one side of the heater circuit can still be connected straight to a cathode (to obtain the heater bias voltage), and there will be no modulation of the cathode voltage by the AC heater supply, because the other side of the heater supply does not reference ground in any way (which would short out the EFB bias if it did!). As well, the connection won't unset the output bias indication, either. So for all practical purposes, the original jumper connection is just fine.
Congrats on getting it going!
Dave
njcanuck
aka: MacGyver
Dave, Gadget : Thanks for the explanation. Helps my understanding considerably. 
Danrclem: The total cost is approx $475 in parts as broken out below.
Dynakit transformers $240
Dynakit chassis $ 95
Dave's EFB & PC-13 boards $ 60
Mouser components. $ 80
Figure another $50 for shipping depending on where you live. Tubes of course depend on what you do or don't already have. Not the cheapest project but well worth it.
Danrclem: The total cost is approx $475 in parts as broken out below.
Dynakit transformers $240
Dynakit chassis $ 95
Dave's EFB & PC-13 boards $ 60
Mouser components. $ 80
Figure another $50 for shipping depending on where you live. Tubes of course depend on what you do or don't already have. Not the cheapest project but well worth it.
Actually a bargain!
If i had kornwalls it might be too much to resist now. (i know these will drive speakers i have already. just making excuses)
I have to enjoy through Rich in the meantime.
If i had kornwalls it might be too much to resist now. (i know these will drive speakers i have already. just making excuses
)I have to enjoy through Rich in the meantime.
danrclem
Super Member
Dave, Gadget : Thanks for the explanation. Helps my understanding considerably.
Danrclem: The total cost is approx $475 in parts as broken out below.
Dynakit transformers $240
Dynakit chassis $ 95
Dave's EFB & PC-13 boards $ 60
Mouser components. $ 80
Figure another $50 for shipping depending on where you live. Tubes of course depend on what you do or don't already have. Not the cheapest project but well worth it.
Thanks for the pricing Rich. I have some 6BQ5 and maybe have some other tubes that would work with it. Oh yeah I've already got a pair of Chorus.
Its really not a bad price. If you were to find an old 6bq5 amp of comparable quality and rebuild it, I'd wager you would be into it for similar money and quite probably more time. Not talking console amps of course, but stuff like a Pilot SA-232, a vintage ST-35, etc.
njcanuck
aka: MacGyver
I was originally going to buy a vintage ST-35 and rebuild it but at going prices, it would indeed be similar in cost. Plus you would be stuck with 8 and 16 ohm taps. While that might work for some people, I have some very good 4 ohm speakers that I want to use. Of course, if you can find a vintage unit for dirt cheap, that changes the equation.
I bet one of those Pilot SA-232s would sound fantastic with Dave's EFB circuit installed but those prices have skyrocketed. There's one on the Bay for $1100 right now and others have sold for $600+. Would love to have one tho . . .
Soon I'll be trying it out on my own Chorus speakers. It should be a very good combo!
I bet one of those Pilot SA-232s would sound fantastic with Dave's EFB circuit installed but those prices have skyrocketed. There's one on the Bay for $1100 right now and others have sold for $600+. Would love to have one tho . . .
Soon I'll be trying it out on my own Chorus speakers. It should be a very good combo!
The 232 has a bit of a non-amazing phase inverter too. Same as the SA-260, and the Magnavox 9300, its a paraphase. I need to drag my own SA-260 back to the bench and finish the floating paraphase conversion that Dave was kind enough to lend me his schematic for. Have to agree about the prices going silly. I got my SA-260 for less than what the 232 is going for now and thats double the power.
I think many of us have wrestled with the concept on these and the ST 70s (probably other pieces too!)
Figuring in possibility it was a kit and miswired or just poorly done, having 40 year old everything, rusting, fading , missing hardware, Not knowing what kind of events it might've suffered through, poor old quality PCBs (which in this case would be gone anyway) Dried out line cord, Transformer (smaller) designed for 117v rather then 120 (necessitating a bucking transformer) and the absurd? (IMO) what people seem to be getting for these old pieces. The route Rich made seems (to me) to be a no brainer!
