In my case I use 2a3 outputs, and I don't have a problem with my 10Y stage with 1:4 step-up.
"In my case I use 2a3 outputs, and I don't have a problem with my 10Y stage with 1:4 step-up."
... at about half value than 300B grid swing. 😉
... at about half value than 300B grid swing. 😉
I guess it would also work using a gyrator, like in the schematics in post #17.
I must admit I don’t understand. Can you please explain? Thank you for the patience..
Practically there are no 1:8 line level step up transformer.
Some manufacturer make for example 1CT:4CT, 1CT:2+2 (1+1:2+2) or 1+1+1+1:2+2+2+2 high primary level transformers.
The manufacturers define frequency response (at a specific connection), usually 600R source impedance requiring, which is mostly not completed in tube equipments.
If you use these transformers with all primary paralleled (or half primary used) and all secondary connected serial (to reach 1:8 transmission) due to the transformer parameters (mostly capacitances) and source impedance, the high frequency response will be degrading.
For example Lundahl defined LL7903 10Hz..70kHz +/- 0.5dB at 600R source impedance and primary and secondary coils connected serially.
Look Ale measurements wired this transformer as 1:8.
https://www.bartola.co.uk/valves/2020/07/15/300b-se-amp-46-driver-part-ii/
It does not mean that this transformer not good, but not developed for such requisitioning.
If you use it as 1:2 or even 1:4, higher bandwidth expected.
Look -my- Sowter 9063 as 1:2
and 1:4 wiring.
Some manufacturer make for example 1CT:4CT, 1CT:2+2 (1+1:2+2) or 1+1+1+1:2+2+2+2 high primary level transformers.
The manufacturers define frequency response (at a specific connection), usually 600R source impedance requiring, which is mostly not completed in tube equipments.
If you use these transformers with all primary paralleled (or half primary used) and all secondary connected serial (to reach 1:8 transmission) due to the transformer parameters (mostly capacitances) and source impedance, the high frequency response will be degrading.
For example Lundahl defined LL7903 10Hz..70kHz +/- 0.5dB at 600R source impedance and primary and secondary coils connected serially.
Look Ale measurements wired this transformer as 1:8.
https://www.bartola.co.uk/valves/2020/07/15/300b-se-amp-46-driver-part-ii/
It does not mean that this transformer not good, but not developed for such requisitioning.
If you use it as 1:2 or even 1:4, higher bandwidth expected.
Look -my- Sowter 9063 as 1:2
and 1:4 wiring.
Last edited:
Ok, so a slight degradation or high frequency roll-off is to be expected. Like you said, no free lunch, yet I guess there is no design without compromises - I need to start somewhere. Thank you for thorough explanation.
Regarding driver/power tube coupling, I plan using a gyrator for now (as I have ordered a bunch of them), however I will deffinitely test an inter stage transformer / choke and cap coupling options if I will not electricute myself or run out of budget sooner.
Regarding driver/power tube coupling, I plan using a gyrator for now (as I have ordered a bunch of them), however I will deffinitely test an inter stage transformer / choke and cap coupling options if I will not electricute myself or run out of budget sooner.
Step-up transformers are often taken for granted to do technically impossible work. Some people unknowingly ask for scenarios like 1:5 up with a 2k Rp SE primary driver. That translates into a reflected Rp at 2x5^2 = 50k. How crazy is that? At such driving impedance, it takes 100pF to screw the frequency response up. Not to mention a potentially high natural secondary leakage inductance resonating with this capacitance. This is why it is realistic to design and use high-ratio step up on low signal, preferably low impedance driving and definitely NO DC current, as it screws the idea of having a small, high-mu core resulting with the fewest number of turns possible to get away from too much leakage inductance.
@euro21 in your experience d3A performs better as a driver if to compare with something with a lower MU + Step up?
What are the drawbacks of using output transformer with lower input impedance, such as monolyth magnetics 3k3?
Triode d3A is very good with gyrator. I like E810F as well. I use C3g triode connected with gyrator load (like Ale does) to drive 2a3. Also, there is E280F. And apparently there are soviet-era valves/tubes that are great as well.
I needed to use ferrite beads on d3A and E810F, but that is not a big deal.
I stopped using inter-stage transformers because:
1. The really good ones that had decent enough impedance with flat response which I liked were... EXPENSIVE.
2. The gyrator loaded triode strapped pentodes have amazing performance. Flat frequency response, low distortion, lots of drive...
3. No, you don't hear the "sand" in the gyrator.
4. Coupling caps have got a LOT better and there is great choice out there. They are cheaper and smaller than inter-stage transformers.
5. Ok, PTFE coupling caps are not cheaper... but still less costly than the inter-stages I liked...
Also, nothing saying you can't do a Gyrator load and use an inter-stage transformer instead of a coupling capacitor. That might be a nice idea in fact if the budget is not an issue... But if you are looking to use the inter-stage to present the load to your input/driver valve/tube, then.. give up right now. It's seldom possible without a significant compromise.
Also, the CCS loaded source follower is a solid performer (and I discovered a fantastic cheap new mosfet recently for this duty too).
Ian
I needed to use ferrite beads on d3A and E810F, but that is not a big deal.
I stopped using inter-stage transformers because:
1. The really good ones that had decent enough impedance with flat response which I liked were... EXPENSIVE.
2. The gyrator loaded triode strapped pentodes have amazing performance. Flat frequency response, low distortion, lots of drive...
3. No, you don't hear the "sand" in the gyrator.
4. Coupling caps have got a LOT better and there is great choice out there. They are cheaper and smaller than inter-stage transformers.
5. Ok, PTFE coupling caps are not cheaper... but still less costly than the inter-stages I liked...
Also, nothing saying you can't do a Gyrator load and use an inter-stage transformer instead of a coupling capacitor. That might be a nice idea in fact if the budget is not an issue... But if you are looking to use the inter-stage to present the load to your input/driver valve/tube, then.. give up right now. It's seldom possible without a significant compromise.
Also, the CCS loaded source follower is a solid performer (and I discovered a fantastic cheap new mosfet recently for this duty too).
Ian
Last edited:
I am planning on using Coleman Fixed Bias regulator for this application - however, it has a max voltage input of 150V. Bartolas designs is asking for 200V for the source follower, however, Coleman guide allows for "any" voltage, as long as their is a suitable Zener (DZ1) accross the input together with an adequately sized (5kΩ 7W) RZDR resistor on the +.
Will this (Zener/Resistor) voltage drom will not introduce noise into my expensively filtered HT supply? Perhaps it would better to form a suitable resistor/capacitor voltage drop from the PSU?
Will this (Zener/Resistor) voltage drom will not introduce noise into my expensively filtered HT supply? Perhaps it would better to form a suitable resistor/capacitor voltage drop from the PSU?
Paulius, the zener on the input of the bias regulator will not generate any noise on the bias output - it will remain below 10µV.
The zener will see a small amount of noise, but this is RC filtered from the 200V supply, and rejected in-circuit because the source-follower has a CCS (high noise rejecting design) to take care of any that remains from RC filtering.
I would be surprised if you can even measure a difference, with or without the zener, but if you want to test it, please add another layer of filtering: use a 4,7kΩ resistor for the dropper, and add 330Ω in series with it. Add a 100µF 200V cap at the node between these resistors. Now try measuring noise with and without this capacitor (and let us know....)
The zener will see a small amount of noise, but this is RC filtered from the 200V supply, and rejected in-circuit because the source-follower has a CCS (high noise rejecting design) to take care of any that remains from RC filtering.
I would be surprised if you can even measure a difference, with or without the zener, but if you want to test it, please add another layer of filtering: use a 4,7kΩ resistor for the dropper, and add 330Ω in series with it. Add a 100µF 200V cap at the node between these resistors. Now try measuring noise with and without this capacitor (and let us know....)
Thank you! So something like this below in the schematics?
It will take some time before I can measure, but surely I will, as its an easy and interesting (well, for me) test to do.
Also, should I add a 1R or 10R Rk resistor to measure the Bias point of 300B in the startup? Its seems to be running very got, I would definitely try to start up up with a strong negative bias...
Paulius
It will take some time before I can measure, but surely I will, as its an easy and interesting (well, for me) test to do.
Also, should I add a 1R or 10R Rk resistor to measure the Bias point of 300B in the startup? Its seems to be running very got, I would definitely try to start up up with a strong negative bias...
Paulius
Paulius, Please connect the POSITIVE input & output of the Bias Regulator to Ground.
Then, the RZDR resistors are inserted in the Negative supply line, as per Post #30.
Yes, 1Ω current-sense resistor in the DHT cathode line can measure the cathode current. It's a good idea to measure it - some driver circuits can send positive-pulses at start-up.
Then, the RZDR resistors are inserted in the Negative supply line, as per Post #30.
Yes, 1Ω current-sense resistor in the DHT cathode line can measure the cathode current. It's a good idea to measure it - some driver circuits can send positive-pulses at start-up.
Hi,
Just doing some tests on the driver stage assembly and noticed, that I have made some mistake in components selection - I cannot set the Gyrator Anode output voltage higher than ~210V. Input side I have got enough headroom, both boards are operating the same way, I guess that must be component selection related. Perhaps somebody has experience using them and could help me? Trying to reach a 15mA 265V operating point for VT-25 with 1kOhm bypassed Auto bias.
I am using:
Just doing some tests on the driver stage assembly and noticed, that I have made some mistake in components selection - I cannot set the Gyrator Anode output voltage higher than ~210V. Input side I have got enough headroom, both boards are operating the same way, I guess that must be component selection related. Perhaps somebody has experience using them and could help me? Trying to reach a 15mA 265V operating point for VT-25 with 1kOhm bypassed Auto bias.
I am using:
CCS (M1, M2) generates required current. This current on R4 produces VREF, which is the "gyrator" reference voltage.
If I(ccs)*R4 is too low, this will be the limiting factor.
Required reference voltage range would be about 250..270V, so with R4=390K the required CCS current is:
270V/390k= 0.69mA
--------------
The CCS HV must be -about- 30V higher, than voltage on R4, so at least 270V+30V= 300V.
----------------
If you want to us this device as driver, the gyrator B+ must be:
265V (+ Rmu*Ia) + 30V (gyrator spare voltage) + required swing peek voltage/2.
--------------
Try to test gyrator with dummy load:
265V/0.015A= 17 667 Ohm ... 17.5 or 18k 10-12W resistor.
If gyrator fulfil the required parameters, it' ready to operate.
If I(ccs)*R4 is too low, this will be the limiting factor.
Required reference voltage range would be about 250..270V, so with R4=390K the required CCS current is:
270V/390k= 0.69mA
--------------
The CCS HV must be -about- 30V higher, than voltage on R4, so at least 270V+30V= 300V.
----------------
If you want to us this device as driver, the gyrator B+ must be:
265V (+ Rmu*Ia) + 30V (gyrator spare voltage) + required swing peek voltage/2.
--------------
Try to test gyrator with dummy load:
265V/0.015A= 17 667 Ohm ... 17.5 or 18k 10-12W resistor.
If gyrator fulfil the required parameters, it' ready to operate.
Thanks! I expect arround 550kOhm for desired result with some headroom, will try with 510k I have got on hand. I have got +380 on the B+ input
Thanks for the tip, now I can set 265V on VT-25.
That was one big breadboarding lesson..
That was one big breadboarding lesson..