I learned about the FP230 series of transformers from Triad. These are flat pack, 2 primaries and 2 secondaries wound on a 4 chamber bobbin, so with guaranteed insulation between the windings. The 230 refers to the secondaries: 2x 115V. I used these as filter chokes in two Futterman OTL amps, where they perform very well.
https://www.mouser.ch/c/power/transformers/power-transformers/?q=fp230
Another application is as a center tapped PP choke. Due to higher RDC at secondaries, I connected the transformer as shown. The resistors and pots are to increase the voltage drop, which is used for biasing the output tubes. The amp uses an input transformer, I choose the cheap TY-250P, also from Triad. The cathode resistor is 154R. I choose it over LEDs or CCS as it helps in keeping the current through the output tubes stable with varying mains AC (tested with a Variac). The actual B- is rather - -190V DC. For the output stage, the 4P1Ls were wired as triode, and fed from a VCCS.
The power supply uses a PT from a defect SS amplifier, which puts out about 88VAC. This goes in the doubler, and with a level shifting cap, to form the negative supply, with some additional filtering.
As directly heated, the 4P1L starts drawing current way before the ECC81s are hot and conducting, so I combined the capacitance multiplier with a crude current limiter, based on the 2x 6.2V zeners (had these hand, 12V Zener also ok) and the 2x 10R resistors. When too much current is drawn, the zeners start conducting and the gate of the Fet is pulled down. This works indeed. As a DC coupled amp, it is indeed nice to have a current limiter if something fails during operation.
As output transformers, I used Talema 7VA toroidal power transformers (230V to 7V). Unloaded, they measure about 8V at the secondary, so they present a 6k6 primary at 8R secondary.
The amp works, currently biased at circa 35mA per output tube. 1W at 0.35% THD, about 2.5W max. The frequency response is circa -2dB at 20kHz, and this is solely due to the used OPT. At the plate of the ECC81 the frequency response is flat until at least 30kHz. Amplification is 60x and very low distortion, so the PP chokes are doing their job.
I tested current disbalance in the FP230-25, and it can take some mA without getting much worse (I will get the exact numbers). The toroidal OPT were tested with a 5mA disbalance (so one side pulling 5mA more than the other). The THD doubled (to 0,6% THD at 1W) and the bass response suffered a bit. I will be using these as mid/high amps with a 12dB HP at 600Hz, so I do not care much about the bass performance.
There are no caps in the audio signal (not that I think that they hurt, but wanted to try it like this). There is still caps in the PS, so I am considering to feed the OPT through a choke, to isolate the amp from the PS and force class A operation (learned that from Lynn Olson’s website).
What I do specially like is that one can have quite a good performance as PP choke from a 10 USD part. Also, the 6 USD TY-250P surprised me.
Now I still have to listen to it!
https://www.mouser.ch/c/power/transformers/power-transformers/?q=fp230
Another application is as a center tapped PP choke. Due to higher RDC at secondaries, I connected the transformer as shown. The resistors and pots are to increase the voltage drop, which is used for biasing the output tubes. The amp uses an input transformer, I choose the cheap TY-250P, also from Triad. The cathode resistor is 154R. I choose it over LEDs or CCS as it helps in keeping the current through the output tubes stable with varying mains AC (tested with a Variac). The actual B- is rather - -190V DC. For the output stage, the 4P1Ls were wired as triode, and fed from a VCCS.
The power supply uses a PT from a defect SS amplifier, which puts out about 88VAC. This goes in the doubler, and with a level shifting cap, to form the negative supply, with some additional filtering.
As directly heated, the 4P1L starts drawing current way before the ECC81s are hot and conducting, so I combined the capacitance multiplier with a crude current limiter, based on the 2x 6.2V zeners (had these hand, 12V Zener also ok) and the 2x 10R resistors. When too much current is drawn, the zeners start conducting and the gate of the Fet is pulled down. This works indeed. As a DC coupled amp, it is indeed nice to have a current limiter if something fails during operation.
As output transformers, I used Talema 7VA toroidal power transformers (230V to 7V). Unloaded, they measure about 8V at the secondary, so they present a 6k6 primary at 8R secondary.
The amp works, currently biased at circa 35mA per output tube. 1W at 0.35% THD, about 2.5W max. The frequency response is circa -2dB at 20kHz, and this is solely due to the used OPT. At the plate of the ECC81 the frequency response is flat until at least 30kHz. Amplification is 60x and very low distortion, so the PP chokes are doing their job.
I tested current disbalance in the FP230-25, and it can take some mA without getting much worse (I will get the exact numbers). The toroidal OPT were tested with a 5mA disbalance (so one side pulling 5mA more than the other). The THD doubled (to 0,6% THD at 1W) and the bass response suffered a bit. I will be using these as mid/high amps with a 12dB HP at 600Hz, so I do not care much about the bass performance.
There are no caps in the audio signal (not that I think that they hurt, but wanted to try it like this). There is still caps in the PS, so I am considering to feed the OPT through a choke, to isolate the amp from the PS and force class A operation (learned that from Lynn Olson’s website).
What I do specially like is that one can have quite a good performance as PP choke from a 10 USD part. Also, the 6 USD TY-250P surprised me.
Now I still have to listen to it!
Attachments
I have not tested the inductance, but I read other reports that it is very high, over 100H. Easiest to measure is putting a resistor to measure current when powered by 230V AC. This assumes a perfect DC cancelation from the PP output stage, that requires some special care in the amp session.
Each winding is made to work at 115V AC 50Hz, and I assume there is little margin there. At 20Hz it will be able to take 20/50*115VAC = 46VAC, which is not a lot... Other builders did therefore put 2 110V secondaries in series, so the transformer works better at lower frequencies. I am aiming at low power and frequency response from 600Hz onwards, so the 115V would be ok for me. But these OPTs drop -2dB at 20kHz. Using feedback, this could be improved, but I don't know if I will go for a second experiment with these. I like cheap, but maybe this is too much of a compromise?
best regards,
Erik
Each winding is made to work at 115V AC 50Hz, and I assume there is little margin there. At 20Hz it will be able to take 20/50*115VAC = 46VAC, which is not a lot... Other builders did therefore put 2 110V secondaries in series, so the transformer works better at lower frequencies. I am aiming at low power and frequency response from 600Hz onwards, so the 115V would be ok for me. But these OPTs drop -2dB at 20kHz. Using feedback, this could be improved, but I don't know if I will go for a second experiment with these. I like cheap, but maybe this is too much of a compromise?
best regards,
Erik
Hi Hooman. Long ago member Shoog did publish a lot of experiences with toroidal PT, even using them as input transformers in amps. Lately member @kodabmx has pusblished positive results with toroidal PTs, see the link as an example.
https://www.diyaudio.com/community/...ut-i-have-4-ohm-speakers.391354/#post-7151020
In another group I got the reply from a person that has developed some toroidal OPTs. Besides mentioning that they are very complex, he told that using standard toroids can work well in class A, but as soon as the amp enters class B (one side cut off) one can get nasty THD response as the transformer is not loaded in a balanced way anymore. So he recommends staying in class A. As DC disbalance is also bad for toroidals, it may be indeed be best to go for an autobias solution with eg LM317 as CCS (bypassed of course) in the cathode, limiting operation to class A and ensuring DC balance as tubes age.
https://www.diyaudio.com/community/...ut-i-have-4-ohm-speakers.391354/#post-7151020
In another group I got the reply from a person that has developed some toroidal OPTs. Besides mentioning that they are very complex, he told that using standard toroids can work well in class A, but as soon as the amp enters class B (one side cut off) one can get nasty THD response as the transformer is not loaded in a balanced way anymore. So he recommends staying in class A. As DC disbalance is also bad for toroidals, it may be indeed be best to go for an autobias solution with eg LM317 as CCS (bypassed of course) in the cathode, limiting operation to class A and ensuring DC balance as tubes age.
Today I put this amp to play and noticed a hummmmmmmmmm. Measured it, a mix of 50Hz and 100Hz and all of it addictions and subtractions. I played with the filtering of B-, referenced the ECC81 heaters to 40V relative to B-, nothing helped really. Then I replaced the cheap input TX with LL1540, which are in a nice case of mu metal, and all the noise went about 15dB lower. So in the end it was the input TX that was picking up "stuff" from the workbench. Lets put it in a mu metal case as well?
