Help fusing valve rectifier preamp supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dear all,

I am rebuilding my preamp and have been reading up on fusing. Previously only the primary side was fused, and now I feel it is good practice to fuse the secondary as well.

Could someone more experienced tell me if my logic is correct here, and if not correct it?

The preamp is simply an ECC82 and ECC83 RIAA stage with a 5687 line output. I've estimated a 50mA draw for these three valves. The secondary is 700VCT, which PSUD2 wants half of. The Hammond datasheet (373X transformer) says 357V. The peak inrush current I get is 245mA in the first choke.

Attached is a screenshot of the supply in PSUD2. I haven't used it before, so I could well have got it wrong.

Should I thus have a 250mA T fuse just before the first choke?
 

Attachments

  • PSU.PNG
    PSU.PNG
    4.4 KB · Views: 136
I can suggest the linked article to help identify a likely suitable fuse rating. The article uses PSUD2 to simulate the time and rms current levels used to confirm that a fuse will not blow, and to help identify the lowest rating fuse that should do the job.

Did you measure the transformer primary and secondary resistances, and choke resistances yourself?

https://www.dalmura.com.au/static/Valve%20amp%20fusing.pdf
 
It is common practise to place an HT fuse in the centre tap to ground line on that type of full wave supply. That avoids DC sparking if the fuse fails due to excess current.
A slow blow or time delay 200mA rated at 250v will be ideal as that value will protect the valves but 250mA is acceptable as the valves will stand 440mA peak current.
C1 should be 32uF.
 
trobbins - funnily enough I posted this after reading that whole document, along with trawling here for previous posts, where I first found it! It was a bit rarefied for me and didn't cover just how to deal with valve rectifiers in a way I understood. Figure 5 had me racking my brains a bit.

I attach the pre-amp PSU diagram I'm working from. This is what I had in the amplifier before. Am I right in thinking that the 5Y3WGTA is directly heated, so the centre tap of the 5V heaters is actually 5V floating on top of the B+?

JonSnell - Is the 32uF you mention a result of the particular valve's tolerance for inrush current? The 100uF is there because of the dual capacitor that was supplied to me when I first made the pre-amp, perhaps because it was originally rectified with a 6X4.

Could you clarify what you mean by 'centre tap to ground line' please? Isn't the fuse just in-line with the yellow/black before the first choke?

Do you prefer the 200mA T fuse because its failure point is closer to the calculated 245mA if a fault current were to be flowing, and a 250mA would have to have more current flowing or a longer time period before failure, which could be at a level that would damage the valves? Does that mean my method of arriving at a fuse value was right-headed?

DF96 - My understanding is that the primary side fuse won't protect the amplifier or the transformer from a fault on the secondary side, as it is likely to blow after the damage has been done. By having a fuse in the B+ I am protecting both the valves and the transformer from fault currents. Is that correct?

Thanks for your time.
 

Attachments

  • Bluebell PSUv2.jpg
    Bluebell PSUv2.jpg
    447.1 KB · Views: 104
What sort of 'fault current'? Are you expecting something in the amplifier to suddenly put a short across the supply rail? If so, then a smaller fuse could be at the PSU output. If you fear the PSU capacitors developing a short then the fuse needs to be where you have suggested. If you fear the rectifiers shorting then you may need two fuses, one in each secondary leg.
 

PRR

Member
Joined 2003
Paid Member
...ECC82 and ECC83 RIAA stage with a 5687 line output. I've estimated a 50mA draw ...

Seems high. Show your work.

When adding fuses you always have to think about what are you protecting, and what are you protecting it from?

Absolutely!!

Then you also have to know the fault current for every fault, at the point you propose a fuse.

Let us assume a 100 Watt transformer. Probably 90% efficient. At normal full load this is around 90W in load and 10W heat in the transformer.

Put a dead short on it. The 10% losses suggest it will suck 1,000W, all of that dissipated in the transformer, designed for 10W losses. It will heat-up fast (but many seconds).

If we fuse the primary for max 200W, a 1,000W demand will blow a fuse very fast. And by ratio of size of transformer to size of fuse, probably the fuse vaporizes before the transformer starts to toast. A slightly generous primary fuse DOES protect the transformer against a short. (That's how utility power systems work.)

Actually a PT-burn is a survivable disaster. If the Line Cord burns it *could* set fire to the carpet and thus burn the house. In yesterday's news a barn fire killed a goat, but you may have more valuable critters near your hi-fi.

Let's go to the small end. A gain triode with typical 100K plate resistor CAN'T suck a lot of power. Worst that can happen, usually, is internal tube short puts full B+ across that 100K. In high-abuse amplifiers, we figure say 0.3W normal resistor dissipation and use a 2W part, so changing the tube puts the show back on stage. The over-size resistor is far cheaper than finding a 2mA fuse.
 
If your PSUD2 circuit values are a good estimate of what you have, then a fuse in the CT has to pass a steady-state 51mArms (PSUD2 shows that as the first choke rms current when using a long reporting delay time). So the first query is would you use a UL type fuse or IEC type fuse, which depends on your supplier.

If you can get IEC type fuses, then the simplest closest rating that is equal to or above 51mA would be 80mA or 100mA. Let's assume you can get a 100mA IEC fuse.

If you set the PSUD2 reporting delay to 0 sec, and the simulate time for 20ms, the choke current has a 100mArms level during that initial period. Similarly for 150ms it is nearly 200mA, and for 600ms it is 133mA. Compared to a 100mA fuse rating, those simulated current levels are 1.0, 2.0, and 1.33 times the fuse rating. Those time periods and multiplier levels indicate a 100mA 'T' fuse will easily survive turn-on.

If you set the PSUD2 reporting delay to 0 sec, and the simulate time for 10ms, the choke current has a 60mArms level during that initial period. Similarly for 50ms it is 164mA. Compared to a 100mA fuse rating, those simulated current levels are 0.6 and 1.64 times the fuse rating. Those time periods and multiplier levels indicate a 100mA 'F' fuse will easily survive turn-on.

If you short the second cap, PSUD2 shows the choke resistance limits the fuse current to about 335mArms, which is a 3.3 times multiplier of a 100mA fuse rating, indicating a 100mA F fuse should blow in less than 2 secs, and probably in under a few hundred milliseconds.

If you short the output then the max current due to choke resistances is about 200mA - which may not blow a 100mA F fuse for some time. And those shorts relate to 'bolted' shorts, which may not be the case if a valve shorts internally.

You should use a 1N4007 in series with each 5Y3 anode to alleviate the risk that a 5Y3 fault takes out the power transformer.

So I reckon if you can get a 100mA IEC F fuse then that would provide some peace of mind, although an 80mA fuse would be better (need to double check the multipliers, but that should also be fine). The caveat is that you have adequately measured the transformer resistances and other part values for the simulation.

PS. If you lower the choke values for the short circuit simulations, you get a bit higher prospective fuse current, which is more likely to be the case, but depends on the choke performance at higher current.
 
Last edited:
Thank you all for the interesting reading. I’ll make clear that this is one of my many hobbies, so while I’m confident in my safe building skills, with my grandfather now deceased I have no ready guru when I’m confused about valve circuit theory and calculation. There is only so far one can go with books and a physics A-level when one’s specific question is not answered satisfactorily in print! Thus the delight of DIYAudio, with the hive-mind's ability to take the place of the knowledgeable friend when no-one you know has any idea about electricity. Already with these answers it’s clearer how to go about conceiving what’s going on here.

DF96 – from the article, I thought the author was referring in particular to faults arising when the biasing on power valves went south. I can see this might cause problems if the secondary was not fused. While this isn’t a concern with this pre-amp as the operating current is not so high, I imagined a damaged resistor could cause out-of-spec current to flow through the valves.

It may be that this fusing idea is over-cautious for this amplifier, in which case it is educational to the inexperienced like myself to have it explained step by step! In my car, I would fuse to protect the wiring loom from overheating, but here I am imagining that the fuse would protect both the transformer and the valves. Perhaps this is muddle-headed.

PRR – this was ‘back of the envelope’ as I was considering the fuse concept, based on operating points cribbed from datasheets, my thinking being that a few milliamps difference from my real operating points wouldn’t make much odds to the usefulness of the idea before I was ready to implement it. As these are dual valves, I doubled it for an overall current draw for each channel.

That worked example is very lucid, thank you. It really helps to have things put down in this manner – knowing what failure states to consider for each part and how to follow the consequences through is not explained so succinctly in any of my books.

Alan4411 – This is another useful mind-jogger – of course a bulb can be used in much the same way as a fuse! I see from PRR’s last point that equally a resistor of low dissipation would burn out in an over-current state, though this would be more difficult to repair than a bulb...

trobbins – I am still working through your points with PSUD2 at my elbow, so to avoid this post becoming a thesis I’ll save my response for when I understand it!
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.