Hi
Over the last 2-3 years I´ve been collecting RCA 808 tubes in order to build a good looking class A2 SE DHT amp.
I´ve already written a lot about this project in other threads on this forum so let´s get straight to the topic: Heatings those hungry filaments in a hum-free and cost effective way.
The filaments draws 7,5V 4A each and AC heating is out of the question since µ=49. I have already built a pair of old-fashioned LCLC-filtered power supplies but I´m not satisfied with them so let´s forget about them.
I recently bought a 12V 10A SMPS on Ebay in order to warm up a bunch of IDH tubes in a future OTL project. As this kind of PSUs can be had almost for free from various chinese manufacturers I also bought two 12V 8,5A units from another seller.
The plan is to use one SMPS per channel to feed the filaments of those 808s through adjustable mosfet-based constant current sources.
RC-filters or voltage regulators would perhaps be easier ways to drop the voltage from 12V (actually 10,8V as the power supplies are adjustable +-10%) to 7,5V, but according to several sources constant current is better than constant voltage in this context.
A couple of questions:
1: Should I expect trouble with noise and RF interference?
2: What will happen if I decide to use cathode feedback in the output stages? I assume that the SMPSs are insulated from input to output, but does that mean that it would be safe to let the output swing, say, +-40Vp WRT ground 20Hz-20kHz?
Here´s the SMPS in question:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=300184452328&ssPageName=STRK:MEWN:IT&ih=020
Over the last 2-3 years I´ve been collecting RCA 808 tubes in order to build a good looking class A2 SE DHT amp.
I´ve already written a lot about this project in other threads on this forum so let´s get straight to the topic: Heatings those hungry filaments in a hum-free and cost effective way.
The filaments draws 7,5V 4A each and AC heating is out of the question since µ=49. I have already built a pair of old-fashioned LCLC-filtered power supplies but I´m not satisfied with them so let´s forget about them.
I recently bought a 12V 10A SMPS on Ebay in order to warm up a bunch of IDH tubes in a future OTL project. As this kind of PSUs can be had almost for free from various chinese manufacturers I also bought two 12V 8,5A units from another seller.
The plan is to use one SMPS per channel to feed the filaments of those 808s through adjustable mosfet-based constant current sources.
RC-filters or voltage regulators would perhaps be easier ways to drop the voltage from 12V (actually 10,8V as the power supplies are adjustable +-10%) to 7,5V, but according to several sources constant current is better than constant voltage in this context.
A couple of questions:
1: Should I expect trouble with noise and RF interference?
2: What will happen if I decide to use cathode feedback in the output stages? I assume that the SMPSs are insulated from input to output, but does that mean that it would be safe to let the output swing, say, +-40Vp WRT ground 20Hz-20kHz?
Here´s the SMPS in question:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=300184452328&ssPageName=STRK:MEWN:IT&ih=020
I performed some measurements on SMPSs for heating a 300B. Maybe they are of interest to you:
DC heater supply part 3, the final insult
I tested just a couple of SMPSs, but it seems that an SMPS is only really clean if the output is grounded. This is fine if the cathode is connected to ground but not with autobias.
Why were you not satisfied? (just out of curiosity)
DC heater supply part 3, the final insult
I tested just a couple of SMPSs, but it seems that an SMPS is only really clean if the output is grounded. This is fine if the cathode is connected to ground but not with autobias.
Why were you not satisfied? (just out of curiosity)
Thanks for your replies.
If necessary I´ll ditch the cathode feedback, I´m not sure it would work properly in a class A2 circuit anyway.
Would common mode between the SMPSs and the CCSs improve anything?
My LCLC supplies works OK electrically but construction-wise they are a mess. I could rebuild them of course, which is my plan B if the SMPS - CCS idea fails.
If necessary I´ll ditch the cathode feedback, I´m not sure it would work properly in a class A2 circuit anyway.
Would common mode between the SMPSs and the CCSs improve anything?
My LCLC supplies works OK electrically but construction-wise they are a mess. I could rebuild them of course, which is my plan B if the SMPS - CCS idea fails.
I have plan to supply 10V 10AM SMPS + 2x in parallel LTC3780.
I was wondering wondering if that would present some problems of oscillations between SMPS and ltc3780(SMPS operate on 1 f and ltc3780 on another.. If same can be problem?)_ hope I explain enough good.
What is with impedance in configuration like that??
FULING: ER du ferdig med 808? Hvis ja jeg vil ha schematic's takk
I was wondering wondering if that would present some problems of oscillations between SMPS and ltc3780(SMPS operate on 1 f and ltc3780 on another.. If same can be problem?)_ hope I explain enough good.
What is with impedance in configuration like that??
FULING: ER du ferdig med 808? Hvis ja jeg vil ha schematic's takk
Biggest problem with mains-powered SMPS is that the way that EMC regulations are complied with.
They do it by adding caps between the output and the mains input (!). This is the reason that you measure a leakage current from the output to mains-ground. It reaches into the milliamp region, so it can't be ignored - unless the output is directly grounded.
So this current flows into the filament, through the cathode resistor, and out to ground.
The leakage current is very noisy, naturally.
Medical versions of SMPS do the job properly, and leakage current drops to microamps - but cost increases by 30-40dB !
They do it by adding caps between the output and the mains input (!). This is the reason that you measure a leakage current from the output to mains-ground. It reaches into the milliamp region, so it can't be ignored - unless the output is directly grounded.
So this current flows into the filament, through the cathode resistor, and out to ground.
The leakage current is very noisy, naturally.
Medical versions of SMPS do the job properly, and leakage current drops to microamps - but cost increases by 30-40dB !
Thanks for replay. Mr. Rod.
Here is LTC3780 which I would put after smps
http://cds.linear.com/docs/en/datasheet/3780ff.pdf
For filament 10V 10A (833A).. These will be future project -hope these year. Mybe you will have something for me to.? which redirect me on second subject (project).
I was just reading 3-4 different post about your regulators. The last question which I have can I use your reg for Stephie Bench LR no C design.... his power supply is ...nja...nja..nja nja ....OK but there is room for improvment. here is link to download his design.
Konzept und Text: Christian Rintelen - Kommunikationsstrategien, Werbekonzepte, Werbetexte, Audio-/Video Werbung, Adaptationen und Übersetzungen, Journalistische und PR-Texte
Here is LTC3780 which I would put after smps
http://cds.linear.com/docs/en/datasheet/3780ff.pdf
For filament 10V 10A (833A).. These will be future project -hope these year. Mybe you will have something for me to.? which redirect me on second subject (project).
I was just reading 3-4 different post about your regulators. The last question which I have can I use your reg for Stephie Bench LR no C design.... his power supply is ...nja...nja..nja nja ....OK but there is room for improvment. here is link to download his design.
Konzept und Text: Christian Rintelen - Kommunikationsstrategien, Werbekonzepte, Werbetexte, Audio-/Video Werbung, Adaptationen und Übersetzungen, Journalistische und PR-Texte
It's OK to use without the 317: they will only waste 3V or more of voltage drop.
The Coleman Regulators will run from rectified dc, and still give ripple-current in the low microamp region. No hum at all, even in a preamp, provided the B+ is well implemented, too.
Tom's regulators are super efficient and great performers, noise-wise. I use them on the 841s in my Unity-Coupled amp in the blog in my sig line. They run cool and give my unobtanium tube filaments a soft-start to help them have a long life. Highly recommended. No heatsink necessary whatsoever.
I was going to design something similar but Tom already made it and did a better job than I would have.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.