Hello DIYers around the World,
I am planning this project for a looong time now, but the output transformer was the problematic part. I am working on the PCBs and the modernized, simplified 313-A output transformer and of course the version with octal pentodes 1620s or 6J7s. I invested in a shorted 313-A OT that was sitting forever on ebay. Unfortunately, despite managing the impossible and cutting through two (probably mumetal) cans and removing the tar without breaking the leads and even removing the bobbin from the core, I couldn't unwind it successfully. But, I've managed to measure it using analog discovery and get the ratios at least, and I hope from a little help from CJ and other good people we can correct the previous calculations and get closer to actual number of turns or at least good enough ones a modern replacement. As for the mic input transformers, I had good luck with OEP and Vigortronix 003 in many tube and SS projects, so that would be my first choice, but I think there are many options like Cinemag and UTM ones frmo Igor, etc. Before the transformer details, just a note that I would like to share the gerber files for free so everybody can use them with modern PSU for high voltage, heaters (probably regulated DC) and phantom power. If you have any valuable input regarding this circuit and potential issues with building, parts placement and such please share.
The 313-A transformer I've destroyed measures as follows:
It is standard E-I type transformer placed a mu-metal kind of soft iron box filled with black tar (which was surprisingly easy to peel of with bare hands) which is all placed into another soft metal box. I think for the modern version even one box is too much since it was mainly there for strong AM transmitters on every corner in those days.
The laminations are butt stacked (all Es and Is are separated there is no paper gap and the last two Es are upside down to hold the Is and there were two brass clamps with screws to hold the transformer together before it was potted in tar) I will try to recover the pics because my phone screen died and I cannot unlock the old phone currently)
The size of the laminations is as CJ mentioned in one of the old threads, I measured it in mm and it is exactly according to EI-48 metric standard (I am not sure but I think the nominal .625 is in practice exactly the same as EI-48 metric) , so:
48mm length, centre leg is 16mm and the window width and two legs are 0.8mm. The stack height is 18mm.
The DC resistances of the windings:
1-2 (Primary) = 2700 Ohms
3-4 (Half tertiary) = 27 Ohms
5-6 (Half tertiary) = 21.5 Ohms
7-8 (Half secondary) = 184 Ohms
9-10 (Half secondary) = 159 Ohms
I fed the sine wave to primary using AD2 with 5V which measure 3.5156 VAC across the primary and the following on the rest of the taps:
1-2 = 3.5156 VAC
3-4 = 0.083582 VAC
5-6 = 0.083000 VAC
7-8 = 0.33984 VAC
9-10 = 0.34151 VAC
So, according to these measurements (hopefully I didn't screw anything up, I've measured it on two separate days few times and it seems to make sense) We know from the specs that the primary impedance should be 16000 ohms and secondary 150 ohms in parallel and 600 ohms in series which is correct according to these sine wave measurements.
The tertiary windings are 9 ohms in parallel and 36 ohms in series, which is where the 6db difference comes from (so when the tertiary windings are in parallel we have less feedback and more gain and when we have tertiary windings in series we have more feedback and 6db less gain.
Now, I was too excited and rushed the unwinding without taking photos and paying attention to the layers etc. I noticed two (possible three) different enamelled wires but they seems very close gauge, probably as CJ predicted around 0.08-0.09mm thick the most. I was hoping to get to the tertiary layer hoping it was 16AWG, haha, so I can just get the rest of the windings according to it, but the ladies at Langevin sure wound everything with this thin gauge with each perfect layer. I will try to share the photos and double check the gauges exactly.
Looking forward to your feedback (pardon my pun) and I will be winding the prototypes later on and testing them once I finish the PCBs and the prototype.
Open to your ideas and suggestions.
I am planning this project for a looong time now, but the output transformer was the problematic part. I am working on the PCBs and the modernized, simplified 313-A output transformer and of course the version with octal pentodes 1620s or 6J7s. I invested in a shorted 313-A OT that was sitting forever on ebay. Unfortunately, despite managing the impossible and cutting through two (probably mumetal) cans and removing the tar without breaking the leads and even removing the bobbin from the core, I couldn't unwind it successfully. But, I've managed to measure it using analog discovery and get the ratios at least, and I hope from a little help from CJ and other good people we can correct the previous calculations and get closer to actual number of turns or at least good enough ones a modern replacement. As for the mic input transformers, I had good luck with OEP and Vigortronix 003 in many tube and SS projects, so that would be my first choice, but I think there are many options like Cinemag and UTM ones frmo Igor, etc. Before the transformer details, just a note that I would like to share the gerber files for free so everybody can use them with modern PSU for high voltage, heaters (probably regulated DC) and phantom power. If you have any valuable input regarding this circuit and potential issues with building, parts placement and such please share.
The 313-A transformer I've destroyed measures as follows:
It is standard E-I type transformer placed a mu-metal kind of soft iron box filled with black tar (which was surprisingly easy to peel of with bare hands) which is all placed into another soft metal box. I think for the modern version even one box is too much since it was mainly there for strong AM transmitters on every corner in those days.
The laminations are butt stacked (all Es and Is are separated there is no paper gap and the last two Es are upside down to hold the Is and there were two brass clamps with screws to hold the transformer together before it was potted in tar) I will try to recover the pics because my phone screen died and I cannot unlock the old phone currently)
The size of the laminations is as CJ mentioned in one of the old threads, I measured it in mm and it is exactly according to EI-48 metric standard (I am not sure but I think the nominal .625 is in practice exactly the same as EI-48 metric) , so:
48mm length, centre leg is 16mm and the window width and two legs are 0.8mm. The stack height is 18mm.
The DC resistances of the windings:
1-2 (Primary) = 2700 Ohms
3-4 (Half tertiary) = 27 Ohms
5-6 (Half tertiary) = 21.5 Ohms
7-8 (Half secondary) = 184 Ohms
9-10 (Half secondary) = 159 Ohms
I fed the sine wave to primary using AD2 with 5V which measure 3.5156 VAC across the primary and the following on the rest of the taps:
1-2 = 3.5156 VAC
3-4 = 0.083582 VAC
5-6 = 0.083000 VAC
7-8 = 0.33984 VAC
9-10 = 0.34151 VAC
So, according to these measurements (hopefully I didn't screw anything up, I've measured it on two separate days few times and it seems to make sense) We know from the specs that the primary impedance should be 16000 ohms and secondary 150 ohms in parallel and 600 ohms in series which is correct according to these sine wave measurements.
The tertiary windings are 9 ohms in parallel and 36 ohms in series, which is where the 6db difference comes from (so when the tertiary windings are in parallel we have less feedback and more gain and when we have tertiary windings in series we have more feedback and 6db less gain.
Now, I was too excited and rushed the unwinding without taking photos and paying attention to the layers etc. I noticed two (possible three) different enamelled wires but they seems very close gauge, probably as CJ predicted around 0.08-0.09mm thick the most. I was hoping to get to the tertiary layer hoping it was 16AWG, haha, so I can just get the rest of the windings according to it, but the ladies at Langevin sure wound everything with this thin gauge with each perfect layer. I will try to share the photos and double check the gauges exactly.
Looking forward to your feedback (pardon my pun) and I will be winding the prototypes later on and testing them once I finish the PCBs and the prototype.
Open to your ideas and suggestions.
