# SE tube amp OT questions for all the tube amp gurus.

Discussion in 'Glowing Bottle Tube Amp Forum' started by Papa Rufus, Jan 19, 2021.

1. ### Papa RufusTDPRI Member

Posts:
21
Joined:
Nov 15, 2020
Location:
I'm building my 2nd tube amp, a Champ Micro, the schematic and layout taken from Rob Robinettes site. The 1st was a successful build of a 5F1 clone. As I was perusing through the multitude of schematics available on the web I began noticing the many different B+ voltage values being used to feed the output transformer primary on a SE amp.

Curious as to why the huge range of voltages I started searching for more info on this in forums around the web but was only able to find info with regards to turns ratio and OT function.

How critical is the voltage value?
How is the required value determined?
Is the voltage more of a relationship between the B+(or 1) (or 2) and the plate voltage?
Is it merely providing a path for the plate current from the output tube?

2. ### 2L manTele-Meister

Age:
61
Posts:
366
Joined:
Nov 23, 2020
Location:
Finland
Operating voltage and OPT impedance are vafiables which define output power and A-class operates best when operating point comes to the middle of loadline. For guitar amp second harmonics is wanted and then not optimized operation can sound better.

Reply what OPT and power tube are going to use and I can guide you to use Giuseppe Amatos Universal Loadline Calculator. It is excellent tool and after a while it is possible to do it in fraction of the time it takes calculating and drawing loadline to datasheet.

3. ### 2L manTele-Meister

Age:
61
Posts:
366
Joined:
Nov 23, 2020
Location:
Finland
I found Champ Micro data: 370V voltage, tube 12AU7 and OPT 25 000 impedance and it came like this when I biased one tube to 7mA which is almost max power. When there are two triodes parallel impedance what one tube gets is double so I did use 50 000. And when there comes about 16V over cathode resistor I did use 355V because this calculator is based to datasheets but in actual use like when cathode biasing there is a resistor which waste voltage. I just put headroom of 160V which is not yet maxed out which show on low distortions. It looks it would be about 0,6W amp.

Last edited: Jan 19, 2021
Mongo Park likes this.
4. ### NickflFriend of Leo's

Posts:
2,554
Joined:
May 24, 2016
Location:
Florida
Check out @robrob's recent thread where he goes through the process of designing an EF 80 power amp. It'll give you a good overview of the process and why variables are chosen the way they are.

By the way, this sort of thing would fit better in shock brothers or the amp tech center.

5. ### Papa RufusTDPRI Member

Posts:
21
Joined:
Nov 15, 2020
Location:
2L Man, This is precisely what I was after! Although the graphing part and the math behind it is well over my head, this certainly explains the varied voltages i've encountered. I need to up the ante on my education and this shows me the direction I need to go. Thank you for your response.

6. ### 2L manTele-Meister

Age:
61
Posts:
366
Joined:
Nov 23, 2020
Location:
Finland
On that loadline you can see the beauty of bias current which gets "pre stored" to output transformer and voltage peak to peak swing comes much higher than what power supply produce. Also when you select a pentode you see that Triode is not as efficient as pentode when voltage swing can not go nowhere near zero end of loadline.

Just play with that Universal Loadline Calculator. When you select any tube it produces average A-class loadline. Then change impedance, current anode and screen voltages one at a time and watch how Screen voltage effects g1 voltage and how power changes. When you increase headroon you can see its power increase and three different distortions as well and how distortion jumps when you try to get too much headroom out. And when you screw it enough just select tube again.

When you find any amplifier data you can add it there and see immedieatey how its operation looks. When two or more tubes are used parallel just multiple impedance by tube count because load spreads between them. If you design cathode bias circuit then power supply voltage needs to be g1 voltage higher. And cathode resistors comes dividing voltage with current and again if you run multiple tube return current thru one resistor the value needs to be divided by tube count.

7. ### Papa RufusTDPRI Member

Posts:
21
Joined:
Nov 15, 2020
Location:
Thanks for the direction to Rob's EF80 page and please pardon the "newbie" mistake as to proper thread placement.

8. ### Papa RufusTDPRI Member

Posts:
21
Joined:
Nov 15, 2020
Location:
2L Man,
Thanks for all the info!! I'll start attempting to digest it all tonight. Small portions for now I'll start by accessing the Loadline Calculator, looks like an invaluable tool if I'm going to pursue amp building to any degree.

9. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,346
Joined:
Dec 29, 2012
Location:
United States
Actually impedance should be halved. With twice as much plate current you need half the load so a plate load of around 10k would work for a single ended 12AU7 with parallel triodes.

10. ### 2L manTele-Meister

Age:
61
Posts:
366
Joined:
Nov 23, 2020
Location:
Finland
Rob, I wrote how to use that Universal Loadline Calculator for parallel tubes. I did look your schematic and if it was correct one I think there was 25 000 impedance reading. Then in theory two parallel tubes use it equally so when doing tube loadline the impedance needs to be double for it work and thats why I did use 50 000 impedance on that loadline example. And yes, when doing other way to make parallel tubes work like the loadline OPT impedance needs to be halved.

11. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,346
Joined:
Dec 29, 2012
Location:
United States
@2L man, it may be simpler to think in terms of a resistive load. A standard 12AX7 triode is used with 100k plate load and 1.5k cathode. If you parallel the tube for two triodes you double the current through those resistors so you'd need to cut the plate load and cathode resistance in half, 47k and 820 ohms typically, to keep the same bias point and load line. Same principal would apply to impedance. Am I wrong here?

Last edited: Jan 24, 2021
12. ### 2L manTele-Meister

Age:
61
Posts:
366
Joined:
Nov 23, 2020
Location:
Finland
Rob I explained to Papa Rufus now to use the Universal Loadline Calculator which opens from this link:

I did use your Champ Micro as an example which schematic is below because Papa is building it but possibly uses different OPT. Possibly bias is not the same when I did not use cathode resistor value, i just changed current to what looked about OK. I should have done better there because that bias/loadline does not come when cathode resistor is 820 ohms.

I believe you should have used 50 000 ohm impedance when setting loadline and operating point when you designed this parallel tube SE amp which OPT impedance comes 25 000 ohms using 8 ohm loudspeaker and I believe you did. If you didn't please explain what method you did use?

Last edited: Jan 24, 2021
13. ### robrobPoster ExtraordinaireAd Free Member

Posts:
8,346
Joined:
Dec 29, 2012
Location:
United States
The dual parallel triode 12AU7 power amp design came directly from Fender. It's what he used to drive the 22.5K:8 ohm reverb transformer. So unless Leo was way off base 22.5k is good to go.

A typical single ended 6V6 uses a 6.6k to 8k:8 ohm output transformer. Parallel 2x6V6 single ended amps typically use a 3.5k to 4k:8 ohm OT.

Keep in mind that the load isn't that critical. Fender allows a 4 or 8 ohm load (single or both speaker outs used) which would cut the output transformer impedance in half, as in 6.6k to 3.3k in the Deluxe Reverb. Most of us feel that even a one step higher impedance like a 16 ohm speaker connected to the Deluxe Reverb is safe, even with an output transformer impedance of 13.2k.

I tried the load line calculator but I don't see a way to specify cathode bias or a strapped dual 12AU7 triode.

Last edited: Jan 24, 2021
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.