Cavalli Bijou Headphone Amp - Power Supply Issues

[bhjazz]

The Cavalli Bijou headphone amp was designed back in ~2006 and was discussed on Head-fi and Headwize. Unfortunately, Headwize is gone and most builders have moved on from that site. Mine was boxed up for years and I kept chipping away at it until this year when I decided I should get it done.

The design has a 250V power supply which uses an EZ81 rectifier tube and either the 270DAX or 370DAX transformer. Additionally, the ps uses a pair of (unfortunately out of production) IRF820B MOSFETs. The transformer is wired for 120VAC input.

I am starting by only connecting the transformer to mains and the power supply board to the transformer. I am also using a poor man's variac (light bulb version.)

First test: check the transformer/ps board with power and adjust the output to 240-250V with a 2W 100 ohm resistor across the output connections. DMM across the leads of the resistor.

At power up, the output voltage gets to near 20V, then slows, and I begin seeing smoke near, possibly, one of the power supply caps. I quickly cut the power at this point. Back on the bench at inspection there are no burn marks, bulging caps, cracked or burst resistors...nothing. The large power supply caps are new Rubycons, so I do not suspect them.

If the MOSFETs are blown, would this cause something else to overheat on the power supply board? I'm stumped. I checked all on-board resistors, reheated most all of the joints and checked the transformer output power and connections.

Also wondering if the poor man's variac is causing trouble for the MOSFETs. As the current increases I imagine the bulb filament could cause some noise on the line.

Here's a link to my question in the giant Bijou thread out on head-fi. I received some replies from a fellow Bijou builder there, but for the most part the discussion is getting little traction over there. It's a shame.

https://www.head-fi.org/threads/bijou-all-tube-futterman-headphone-amplifier.277541/post-17703421

 

[rayma]

Can you post here the actual schematic?

 

[bhjazz]

The schematic reads "not for reproduction" but it is available on the Internet Wayback Machine, so hopefully Alex Cavalli won't mind.

 

[bhjazz]

If the MOSFETs are blown, would this cause something else to overheat on the power supply board? I'm stumped. I checked all on-board resistors, reheated most all of the joints and checked the transformer output power and connections.

This is becoming a theme for my research. Since I stuffed the boards a year or so back, I can only imagine that the MOSFETs have somehow shorted or received some ill static charge. From what I have read these can fail with either an open circuit or closed and either way there can be too much current going to the wrong place. Eek.

I'm on the hunt to replace the IRF820B MOSFETs.

 

[AllenB]

Thread moved to Power Supplies.

 

[bhjazz]

The biggest contenders (so far) for replacement of the no-longer-manufactured IRF820B MOSFETs have similar parameters. But I don't understand which would be the right replacement. I can get close with the Vds and Vgs, but the Rds seems to be harder to match. Should I continue searching for a better match or use one of the following?

(EDIT: It appears the Rds for the 820B is 2.6 Ohm, not 210 mOhm as sometimes specified. Beware of peculiar datasheet data!)

	IRF820B	2SK2402	IRF820APBF	IRF820APBF-BE3	IRF820ALPBF	STP11NM50N
Package / Case:	TO-220-3	TO-220	TO-220AB-3	TO-220AB-3		TO-220-3
Transistor Polarity:	N-Channel	N-Channel	N-Channel	N-Channel	TO-262-3	TO-220
			1 Channel	1 Channel	N-Channel	N-Channel
Vds - Drain-Source Breakdown Voltage:	500 V	600 V	500 V	500 V	500 V	500 V
Id - Continuous Drain Current:	2.5 A	3.5 A	2.5 A	2.5 A	2.5 A	8.5A (Tc)
Rds On - Drain-Source Resistance:	2.6 Ohm	2.2 Ohm	3 Ohms	3 Ohms	3 Ohm	470mOhm @ 4.5A, 10V
Vgs - Gate-Source Voltage:	- 30 V, + 30 V	- 30 V, + 30 V	- 30 V, + 30 V	- 30 V, + 30 V	- 30 V, + 30 V	±25V
Vgs th - Gate-Source Threshold Voltage:	4.0 V		4.5 V	4.5 V	2 V	4V @ 250µA
Qg - Gate Charge:	14 nC		17 nC	17 nC	2 V	10V
Minimum Operating Temperature:	- 55 C		- 55 C	- 55 C	17 nC
Maximum Operating Temperature:	+ 150 C		+ 150 C	+ 150 C	- 55 C
Pd - Power Dissipation:	49 W	75 W	50 W	50 W	50 W	70W (Tc)
Channel Mode:	Enhancement		Enhancement	Enhancement	Enhancement	-55°C ~ 150°C (TJ)
Packaging:	Tube		Tube	Tube
Brand:	onsemi / Fairchild		Single	Single	Tube
Configuration:	Single		13 ns	13 ns	Single	19 nC @ 10 V
Fall Time:	35 ns		12 ns	12 ns
Rise Time:	30 ns	Toshiba	Vishay Semiconductors	Vishay / Siliconix	Vishay Semiconductors	547 pF @ 50 V
Typical Turn-Off Delay Time:	40 ns		IRF	IRF	IRF	-
Typical Turn-On Delay Time:	10 ns

 

[bhjazz]

The top two contenders to the IRF820B MOSFET seem to be IRF820APBF and IRF820ALPBF. However, the Rds On is higher for both of them and the Vgs th are higher or lower, depending on the part.

I feel like the lower voltage requirement coupled with the higher Rds On resistance would be a safer bet.

As my replacement will likely have an Rds On of 3 Ohm, should I opt for the 2V Vgs th or the 4.5V?

	IRF820B	IRF820APBF	IRF820ALPBF
Rds On - Drain-Source Resistance:	2.6 Ohm	3 Ohms	3 Ohms
Vgs - Gate-Source Voltage:	- 30 V, + 30 V	- 30 V, + 30 V	- 30 V, + 30 V
Vgs th - Gate-Source Threshold Voltage:	4.0 V	4.5 V	2 V

 

[dBel84]

I might have some old stock. Good to see you about and still using the Bijou..dB