Feedback requested on 7591 PP design

tikiroo · 15th September 2008, 02:28 PM

Hi,

I've been gathering all the bits and pieces for my first tube amp build and am looking for some feedback on the design. The basic topology is from Eli Duttman's El Cheapo, but using 7591's in ultralinear mode. Most of my initial questions were answered in this thread, I just need to make sure I've understood everything correctly before I started wiring everything up.

The output transformers that I've got have an 8k primary with 6k taps, using them in ultralinear mode the taps are about 87% (if my maths is correct) so the output valves are essentially operating in triode mode, but maybe giving me a couple of extra watts (I'm guessing about 12W output).

The design is using combination bias, with about 6V over the cathode resistor and -12V applied to the grids.

The schematic is attached, power supply to follow.

tikiroo · 15th September 2008, 02:32 PM

Here is the power supply. The bias supply is using a spare 6.3V winding from each transformer connected in series then voltage multiplied. EZ81's provided a "soft start" for the CCS. Only one is required but I've used 2 because I had them

tikiroo · 15th September 2008, 02:34 PM

Finally a picture of the chassis, with tubes temporarily installed. The front and sides will be hardwood.

kevinkr · 15th September 2008, 04:01 PM

I know Eli is a big fan of combination bias, but I'm not.. I much prefer fixed bias over either cathode or combination bias because of the large, low quality electrolytic required to bypass the cathode resistor.

Not sure what you are trying to accomplish with the EZ81.. Their forward drop is sufficiently high that the ccs connected to them is not going to see much if any B-. You don't need to delay the B- unless the voltage present would exceed the ECC81 filament/cathode breakdown rating which is definitely not the case here.

rdf · 15th September 2008, 04:37 PM

Doesn't combo bias, which uses a smaller value cathode resistor, necessitate even larger cathode bypass caps? Sims and bench results have convinced me cathode caps are as much a factor in overload recovery as grid caps. Direct coupled fixed bias seems the optimum way to go.

Tom Bavis · 15th September 2008, 05:53 PM

Quote:

Originally posted by rdf
Doesn't combo bias, which uses a smaller value cathode resistor, necessitate even larger cathode bypass caps? Sims and bench results have convinced me cathode caps are as much a factor in overload recovery as grid caps. Direct coupled fixed bias seems the optimum way to go.

Then don't drive the amp into overload... simplicity does have its costs.

I agree - DON'T delay the B! If the 7591s warm up first, they will be drawing a LOT of current until the bias comes up.

SHiFTY · 15th September 2008, 08:49 PM

Having built a version of this circuit I can attest that it does sound good. Having said that your power supply is not right.

As it is drawn your B+ voltage willl slam the cold valves with 450V and NO bias voltage, the output valves will conduct at MAXIMUM current possible! The life of your 7591s will be brutal and short.

I think you should use the EZ81s for the B+ instead. They are pretty grunty rectifiers, with a max of 150mA, 50uF reservoir EACH!!! so two in parallel should be ample for the job, with a low forward resistance. You can just sub them right into the circuit you have there. IMHO use diodes and large caps for the B- so it is there immediately.

That way you get a nice slow ramp up for the B+ and correct bias.

Also you might want to get rid of the input cap, I found it unneccessary on my amp.

kevinkr · 15th September 2008, 10:04 PM

Quote:

Originally posted by SHiFTY
Having built a version of this circuit I can attest that it does sound good. Having said that your power supply is not right.

As it is drawn your B+ voltage willl slam the cold valves with 450V and NO bias voltage, the output valves will conduct at MAXIMUM current possible! The life of your 7591s will be brutal and short.

I think you should use the EZ81s for the B+ instead. They are pretty grunty rectifiers, with a max of 150mA, 50uF reservoir EACH!!! so two in parallel should be ample for the job, with a low forward resistance. You can just sub them right into the circuit you have there. IMHO use diodes and large caps for the B- so it is there immediately.

That way you get a nice slow ramp up for the B+ and correct bias.

Also you might want to get rid of the input cap, I found it unneccessary on my amp.

Actually his C- is available within a few cycles of start up, and this provides bias to the output tubes. The B- is delayed or more correctly clobbered by the EZ81s.. I thought about recommending the EZ81s in the power supply, but I remember I had several Eico HF-30 amplifiers with paralleled EZ81s and I figured a pair might be a bit marginal for a stereo amplifier using 4 x 7591 in the same sense as a single 5AR4 is IMHO - it'll work, but it's not optimum.

SHiFTY · 15th September 2008, 11:00 PM

Um, you're right, I need a coffee! B- is usually written as the bias voltage in my experience, didn't see the C- part.

EZ81s are tough valves and would probably be OK for B+ rectification assuming a lowish idle current of 35-40mA or so per 7591, they should still have a large margin of safety. I prefer valve rectification mainly for the slow warm up. Diode rectifiers slam the circuits too hard for my liking.

tikiroo · 15th September 2008, 11:38 PM

Thanks for all the comments guys, just what I was looking for! Most of the circuit details were suggested by Eli, or are a straight copy of the El Cheapo design. There are some compromises due to using parts already at hand.

I'll try to address the points raised:

Use of combination bias - I believe this was to allow the use of higher value grid leak resistors (to avoid loading the ECC81 too much) while still getting some of the extra power available from fixed bias. The datasheet says maximum resistance allowed for fixed bias is 300k vs 1M for cathode bias. Resistance to earth in the present circuit will be about 340k. Fixed bias is definitely an option if I can get away with it. By the way I have a couple of 470 uF Blackgates for the cathode bypass resistors, so a step up from "low quality"!

SS rectification - the voltage of my power transformers is a little low for 7591's if I use tube rectification. I'm aiming for a B+ of close to 400V which I can only achieve with SS rectification.

EZ81's in bias supply - this is copied from the El Cheapo power supply design and I'll admit I wasn't too sure why a delayed B- for the CCS supply was a good thing. I'm also unsure what voltage drop I'll get across them. It would certainly be easier to replace them with another schottky diode and 1000uF capacitor. Perhaps Eli could shed some light on this?

Anyway, I will get started on the rest of the power supply for now.

15th September 2008, 04:01 PM	#4
kevinkr diyAudio Moderator Join Date: Sep 2004 Location: Boston, Massachusetts Blog Entries: 6	I know Eli is a big fan of combination bias, but I'm not.. I much prefer fixed bias over either cathode or combination bias because of the large, low quality electrolytic required to bypass the cathode resistor. Not sure what you are trying to accomplish with the EZ81.. Their forward drop is sufficiently high that the ccs connected to them is not going to see much if any B-. You don't need to delay the B- unless the voltage present would exceed the ECC81 filament/cathode breakdown rating which is definitely not the case here. __________________ "Extraordinary claims require extraordinary evidence." - Carl Sagan

15th September 2008, 04:37 PM	#5
rdf diyAudio Member Join Date: Jun 2004 Location: big smoke	Doesn't combo bias, which uses a smaller value cathode resistor, necessitate even larger cathode bypass caps? Sims and bench results have convinced me cathode caps are as much a factor in overload recovery as grid caps. Direct coupled fixed bias seems the optimum way to go. __________________ Ears aren't microphones.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Help with splitter design for 7591 PP	avincenty	Tubes / Valves	10	28th April 2008 02:35 PM
New to tube building / design - Input requested!	Zap	Tubes / Valves	19	19th November 2007 08:31 AM
FR125 New Design Options - Feedback Requested	ABS	Full Range	74	17th October 2007 07:03 AM
Specifications/Feedback requested FTT75 Ribbon Tweeter with Audio Technology bass/mid	Adrie	Multi-Way	4	20th April 2005 08:07 AM
A roast requested on my TQWT design	Christof	Multi-Way	9	29th March 2004 05:57 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

15th September 2008, 08:49 PM	#7
SHiFTY diyAudio Member Join Date: Jul 2002 Location: New Zealand	Having built a version of this circuit I can attest that it does sound good. Having said that your power supply is not right. As it is drawn your B+ voltage willl slam the cold valves with 450V and NO bias voltage, the output valves will conduct at MAXIMUM current possible! The life of your 7591s will be brutal and short. I think you should use the EZ81s for the B+ instead. They are pretty grunty rectifiers, with a max of 150mA, 50uF reservoir EACH!!! so two in parallel should be ample for the job, with a low forward resistance. You can just sub them right into the circuit you have there. IMHO use diodes and large caps for the B- so it is there immediately. That way you get a nice slow ramp up for the B+ and correct bias. Also you might want to get rid of the input cap, I found it unneccessary on my amp.

15th September 2008, 11:00 PM	#9
SHiFTY diyAudio Member Join Date: Jul 2002 Location: New Zealand	Um, you're right, I need a coffee! B- is usually written as the bias voltage in my experience, didn't see the C- part. EZ81s are tough valves and would probably be OK for B+ rectification assuming a lowish idle current of 35-40mA or so per 7591, they should still have a large margin of safety. I prefer valve rectification mainly for the slow warm up. Diode rectifiers slam the circuits too hard for my liking.

15th September 2008, 11:38 PM	#10
tikiroo diyAudio Member Join Date: Nov 2007 Location: Mount Gambier	Thanks for all the comments guys, just what I was looking for! Most of the circuit details were suggested by Eli, or are a straight copy of the El Cheapo design. There are some compromises due to using parts already at hand. I'll try to address the points raised: Use of combination bias - I believe this was to allow the use of higher value grid leak resistors (to avoid loading the ECC81 too much) while still getting some of the extra power available from fixed bias. The datasheet says maximum resistance allowed for fixed bias is 300k vs 1M for cathode bias. Resistance to earth in the present circuit will be about 340k. Fixed bias is definitely an option if I can get away with it. By the way I have a couple of 470 uF Blackgates for the cathode bypass resistors, so a step up from "low quality"! SS rectification - the voltage of my power transformers is a little low for 7591's if I use tube rectification. I'm aiming for a B+ of close to 400V which I can only achieve with SS rectification. EZ81's in bias supply - this is copied from the El Cheapo power supply design and I'll admit I wasn't too sure why a delayed B- for the CCS supply was a good thing. I'm also unsure what voltage drop I'll get across them. It would certainly be easier to replace them with another schottky diode and 1000uF capacitor. Perhaps Eli could shed some light on this? Anyway, I will get started on the rest of the power supply for now.