Fatman iTube DC Offset Issue - Does Anybody Else Have One?

Jump to Latest
L

LimitedSlippedDisc

Member
Joined 2021
#1
Hi folks,
I’ve recently purchased a little Fatman iTube (silver mk1). I’ve cleaned up the pot and it sounds good to my ears, however the DC offsets at the speaker terminals are -130mv and -140mv which is much higher than I am used to with other amps.

The iTube is a valve pre-amp section (6n1) coupled with a TDA7265 chip amp for the power section.

I have measured the DC at the input pins of the chip and all seems well there. I note on the schematic of the chip that a feedback capacitor is omitted in the data sheet and this appears to be the case on the iTube PCB too. Could this be the issue?

I’m curious if anyone else had one of these and what their offset is. Perhaps it’s just as it was when it left the factory?

Steve
 
#2
Not completely terrible, but a little higher than ideal. Feedback isn't the issue.

The TDA7265 input DC offset voltage spec is 20mV, so even just a gain of seven (17dB) would result
in the 140mV DC output offset. So this is likely to be normal. It should cause no harm though.
 
Last edited:
#3
rayma said:
Not completely terrible, but a little higher than ideal. Feedback isn't the issue.

The TDA7265 input DC offset voltage spec is 20mV, so even just a gain of seven (17dB) would result
in the 140mV DC output offset. So this is likely to be normal. It should cause no harm though.
Click to expand...
Hi Rayma - thanks for the reply. When listening to the amplifier it doesn’t stand out as a problem and the sound quality is very pleasant. But taking that reading has bothered me as I’ve never seen anything more than 50mv before! As a result I took a good look round the PCB to see if anything looked amiss and some of the caps nearest the chip do have some greyish residue, yet no bulging. I think it may be thermal compound. Knowing if this is a typical offset reading for this model would put my mind at ease and let me get on with enjoying it rather than analysing it heh! They are a great looking unit and surprisingly well built and weighty. I did dig up an old post on this forum about the merits of adding a 22uf cap in series with the feedback resistors to reduce offset issues on the 7265, do you think it wouldn’t help?

Cheers, Steve
 
#4
Yes, you could indeed add a large electrolytic capacitor to ground at the bottom of the lower nfb resistor.
It would have to be at least 100uF to work properly. That would reduce the output DC offset to under 20mV.

However, the sound may not be as good with the capacitor. You would also have to determine the value
of the lower nfb resistor in order to properly size the capacitor. So I'd probably leave things as they are.
 
Last edited:
#5
rayma said:
Yes, you could indeed add a large electrolytic capacitor to ground at the bottom of the lower nfb resistor.
It would have to be at least 100uF to work properly. That would reduce the output DC offset to under 20mV.

However, the sound may not be as good with the capacitor. You would also have to determine the value
of the lower nfb resistor in order to properly size the capacitor. So I'd probably leave things as they are.
Click to expand...
Thanks again Rayma. This morning I have taken another look inside and have noted that the output of the chip goes straight to a mini pcb attached to the headphone port and from there wires go to the speaker terminals. So if I understand things right, hooking up headphones will expose them to the same offset and I’d imagine this would potentially damage a delicate voice coil in a headphone driver pretty quickly. Which has made me wonder again if this can be normal behaviour for the amp as buyers would of been frying their headphones all the time. Do you think replacing the resistors and caps with higher accuracy replacements would improve things?
 
#6
No, as in post #2, that offset is expected from the circuit design. Without the 100uF capacitor in the nfb network to ground,
the offset must exist. If you look at the pcb and it seems possible to add the capacitor, that would fix the offset.

Do you have the schematic?
 
#7
rayma said:
No, as in post #2, that offset is expected from the circuit design. Without the 100uF capacitor in the nfb network to ground,
the offset must exist. If you look at the pcb and it seems possible to add the capacitor, that would fix the offset.

Do you have the schematic?
Click to expand...
Hi Rayma, unfortunately I haven’t been able to find a schematic but nfb resistor pairing for each channel is an 18k and a 560r. Looking at the pcb I think I can readily add a capacitor after the 560r to ground. I understand they need to be non-polar, however there isn’t room to use film caps of sufficient value. I’m considering adding Nichicon Muse instead - would 100uf be okay for them please?
 
You must log in or register to reply here.
Top Bottom