DIY Audio Projects Forum

Howdy,

How critical is the B+ voltage? I have a pair of power xformers 325 CT that if I model with the Duncan's amp PSU with a bridge rectifier with a choke input, I can get the B+ down to about 260 VDC without a lot of power resistors. If I get the driver B+ to about 230 VDC is that good enough?

Will this affect the bias as well? (I'm assuming 41mA to about 38mA?)

Thanks.

Hi, That should be fine. With modern EL84s you can be pretty safe up to a 250 volts in this circuit without changing the idle current. On vintage ones I would have some concerns about dissipation. Of course with a CT transformer, you could use it that way and the B+ while a bit lower than design would be sufficient and not cause an audible loss of output.

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:

• PoddWatt: 5751 / EL84 Push-Pull DIY Tube Amplifier Project
• OddBlocks: KT88 Push-Pull DIY Tube Amplifiers (Mono Blocks)
• ForeWatt: ECC802S (12AU7 / ECC82) DIY Tube Preamplifier Project
• GrooveWatt: Tube (Valve) RIAA Phono Preamplifier Project

I've been trying to read about the start up issues with solid state and will stick with what I know. I'm not sure what this will mean for the LR8 and will probably just go for a class A push pull with AC heaters.

I have a pair of 8K 50% ultra linear OPTs for this project. Are they far too off spec wise?

The limitations for the first capacitor after a tube diode are only if it's immediately after the rectifier. Since I have the choke input, the capacitors after can be a bit higher?

Thanks as always.

However, your filter inductor is far too low a value to use with the choke input configuration.

_________________
Matt
It's all about the Glass!
http://www.CascadeTubes.com
Cascade Tubes Blog

Hi, The size of the first capacitor on a solid state supply is really not all that critical as long as it is over about 50uF. The internal impedance of the transformer will be 99.9% good enough to handle the inrush. I find that up to about 330uF will not cause harm to the rectifiers. My normal preference is about 100uF on such supplies. For tube rectifiers you have to limit it to much lower values and it varies with the tube. About the max value is 50uF.

The output transformers you have should be OK. 50% taps are rather scarce now, and the more common ones are in the 40% range. It will reduce the gain a tad and may improve the distortion level about the same. Neither will be audible though.

I tend to agree on the choke value a bit small, but one I might use as they are available and modest cost. I would however reduce the 100uF capacitor to about 50uF and re calculate it be OK. Just a gut feeling on that. The design allows for significant ripple cancellation in both the output and driver stage so while the B+ can't be really dirty, it need not be pristine either. It brings up one of my favorite objections to using PC SIMs in tube circuits (BTW I do use some) in that they may give a solution that doesn't work well. Solid state components usually model well. I find that too often someone will argue that the SIM said it would work (or not work) and argue that the actual circuit is no good. This is where I find experience to come into play. I might run 50 SIMs on a particular gain stage before deciding on one to try. The others would probably work, but the PC is really dumb in saying which is best. Most gain, widest band width or lowest distortion sure....but best sound, no. I get inquiries on some of my designs all the time. Most start out with why didn't I use the "textbook" value for something. Simple, I probably did in a prototype and it was deficient in sound quality. Then it is reconfigure as often as needed to get the sound quality I want. The end value is always tested for the usual stuff (noise, distortion etc) but first listened to. That way the data doesn't get in the way of sound quality. The final results are sometimes surprising.

Worse in some ways is that not all tubes are alike. The spec for a 12AX7 (and the ECC variations) all read pretty much the same. However I have found that there is a huge range of deviations from the spec. That is why I often specify what type and brand tubes I recommend for any particular project. They are selected in advance usually for easy availability, modest cost and consistent performance. Often different ones will do fine, but there have been exceptions particularly with KT88s and 12AX7s. Sometimes the issues are minor, and other times (with some KT88s particularly) can be disastrous.

I'm sure this is more of an answer than you wanted....but I figure it relates to an understanding of how stuff works.

Good listening
Bruce

Good Listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:

• PoddWatt: 5751 / EL84 Push-Pull DIY Tube Amplifier Project
• OddBlocks: KT88 Push-Pull DIY Tube Amplifiers (Mono Blocks)
• ForeWatt: ECC802S (12AU7 / ECC82) DIY Tube Preamplifier Project
• GrooveWatt: Tube (Valve) RIAA Phono Preamplifier Project

Thanks for the nudge...

I've redone the PSU sim with a hammond 193C (20H 181ohms) and changed the first capacitor to 50 uF. I get 233 and 202 VDC for B+ voltages so that is also much better. I measured the PT B+ output with no load and that is where the 341 VAC comes from. If the B+ voltages are too low in real life I can change the power resistors for lower values.

The B+ off the rectifier is 409 VDC before the choke. (Same PT in a similar setup but with 6AN8 - roughly 90mA as well) Is this the B+ I use in determining the minimum choke value? (B+ / total A, in mA)

Thanks to all.

Hello all,

I have been trying to work my way through understanding of this amp before I build it. I think that I have everything down pat except the purpose of the 1meg wire that ties the 205V line to both of the EL84 inputs. I realize there is 2.42Meg of resistance in each case, but it seems like that would put about 10V on each of the EL84 grids when there is no input signal to the amp. Obviously electronics is not my forte. Could someone help me out. Also, while I'm making myself look foolish, why place the 2.2uf and .33PIO caps in that mix too. I have a few very shaky ideas.

Thank you in advance.

Luke

Hi, You are more observant than many folks. A rather astute question. Yes the resistors do put some voltage on the grids. I'll try to simplify the explanation for why. When using a constant current source (CCS) in the cathode circuit the CCS will raise the voltage there by some amount that would equal the bias voltage necessary for the tubes to conduct the amount of current that was established by the CCS. (that's a mouth full) Additionally the signal drive to the tubes can not exceed (negatively) the voltage that the CCS caused to exist. Using real world parts (the LM317) there is an additional concern. The LM317 becomes rather nonlinear as the voltage across it approaches zero. Actually I have seen odd behavior at 2 volts across it. So to compensate..... if you raise the grids by a positive amount it will cause a like increase in the cathode circuit. For example adding 10 volts to the grids will increase the cathode voltage by 10 as well. This increase will make it impossible for the signal to cause the CCS to approach the nonlinear region. The only real consequence of this method is that the CCS will have additional voltage across it and since the current is constant it will have to dissipate more power in the form of heat. There is an effective reduction in the voltage between the anode and cathode in the tubes as well. It is not of consequence in this amp. The overall effect is that the circuit can deliver about 25% more power than without the grid boost and the distortion level at max output is clean and is not contaminated by artifacts caused by nonlinear operation of the CCS. This scheme can be used in other similar amps.....up to the voltage and power dissipation limits of the CCS. So no...this will not work in the EL34, KT77, KT88 and KT120 versions of the amp as the voltage will be too high for this particular IC. Equally in those amps the drive is such that they do not venture into the nonlinear region until after they have already exceeded their max rated outputs. Variations using 6V6, 6L6 and some others (like 6GV8, 6BM8 etc) are feasible.

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:

• PoddWatt: 5751 / EL84 Push-Pull DIY Tube Amplifier Project
• OddBlocks: KT88 Push-Pull DIY Tube Amplifiers (Mono Blocks)
• ForeWatt: ECC802S (12AU7 / ECC82) DIY Tube Preamplifier Project
• GrooveWatt: Tube (Valve) RIAA Phono Preamplifier Project

Bruce,

Thank you for your fast, detailed response. Having these great schematics and forum available to all is pretty wonderful. Having the designer answer your question-- even better! A lot of forums are so full of misinformation that learning correctly is often a circuitous, labrynthian process.

The CCS makes more sense now, and the capacitors' purposes snap into place as well. Once I work through the power supply, I'll start ordering parts. I'll try to post a photo when completed.

Thanks again for saving me a whole lot of sleep.

Luke

Bruce,

It's been a few months but I thought that I would post some photos. My experience with tube technology consists of building a Hi-Watt clone guitar amp with very detailed directions and I have never listened to any home "stereo" tube amps. Needless to say I was blown away by how nice this amp sounds. Strangely I don't listen to music much, but I find myself turning this on just to hear it. Very smooth!

tube purists skip to the pictures...

I integrated an Arduino Mega 2560 microcontroller to do various things. Also, some did not work out as planned.
--switches inputs and power to the tube amp via 5VDC relays. I kept everything stereo until an Edcor Stereo-to-mono converter just before the volume pot. I did this in case I added another mono block down the road.
--Acts as an onboard MP3 (random) player. I actually had to add a second simpler Arduino Uno to accomplish this due to interactions with software.
--Powers a bluetooth audio receiver. Sounds great!
--Runs an extremely accurate clock, and displays the time as well as input selection on a LED array. The brightness of the LED panel follows the rooms light levels by use of a phototransistor.
--Controls a cooling fan by using a thermistor to monitor temps around power section.
--Provides user interface through touch senstive capacitance sensors (aka brass rod) inlaid in the front panel.
--Receives and decodes IR from an Apple tv remote. All of the functions are available both by remote and touch sensors.
____Volume up and down, by using a Bourns motorized pot.
____input selection MP3, Bluetooth, 3.5mm, RCA, FM (despite several tries FM is not functional. Yet....
____Mute (also pauses MP3 while in mute mode)
____Power to tube amp
____Next random MP3 song(Next/previous FM preset if I can get an FM to work, previous MP3 track not figured out yet)

Arduino sketch and components list available for the asking. The learning curve on the code writing is rather steep but manageable even for a beginner. You can get whatever level of feedback you like from your computer as you need it for troubleshooting and it runs trouble free on its own when debugged. I highly recommend it. It seems like you would be able to just copy and paste your way to a great project but it actually involves a lot of critical thinking and loads of tinkering. I learned a lot about circuits of all kinds and built a lot of confidence in electronics. It was a bit of a revelation to get such good, instantaneous feedback from every circuit I worked on.

https://www.flickr.com/photos/lukesukes/16265948317/
The inlaid wood box holds 5VDC and 9VDC switching power wall warts. I've added a lot of filtering to the 5VDC during a bout of noise chasing (in hindsight, not needed) and shielded the box's interior with the paint on stuff. The amp is much quieter than I expected or deserved.
https://www.flickr.com/photos/lukesukes/16450901142/
https://www.flickr.com/photos/lukesukes/16265941047/
Despite enlarging the chassis to what initially seemed like gigantic proportions based on the decriptions and photos on the website I ended up rather cramped due to all of the arduino gadgetry. I had to do a good bit of troubleshooting of both the amp and arduino components after final assembly and fiercely wish I had more space. Live and learn.
https://www.flickr.com/photos/lukesukes/16451846705/

Thank you for the encouragement and the great plans. What a great amp!

Luke