Post 2277 above says the input should be as low as possible. I can get the output down to zero with 20mv on the output. So now quite confusedinput orcoutput minimum. Can someone clarify? Thanks
Oops,
That should have read I am now confused as to whether the input or the output of the AD844 should be set for zero offset.
I would be extremely grateful if someone can please advise on the best way to trim the offset.
Thanks
Ian
That should have read I am now confused as to whether the input or the output of the AD844 should be set for zero offset.
I would be extremely grateful if someone can please advise on the best way to trim the offset.
Thanks
Ian
Looking back through the thread I am pretty sure that the offset should be adjusted to zero offset on the input of the AD844 and therefore my offset of 400mv on the output is very high. So I am again thinking about subtle adjustment to the positive or negative supply to AD844 to reduce the offset . Does this approach make sense?
Hi,
I need some advice to the Pedja Rogic AD844 stage (see schematics e.g. in post #7 of D3a IV stage)
I assembled it with 2 ADs stacked (output pins 6 connected through 68 Ohm resistors). But I'm not able to nullify the pin 5 offset and the voltage seems way off. I suspect my ADs are fakes. I bought them from China ebay because I have no other source for DIP package.
Anyway the DC voltages are: pin 1...-15.04, 2... 0.45, 3.... 0 (gnd), 4... -15.04 (supply), 5... -11.7, 6... -12.9, 7... 15.04 (supply), 8... 0
The JFET null circuit has little influence on pin 5, some 0.2 V, but no way to go to zero. Both channels are basically same. What do you think?
I need some advice to the Pedja Rogic AD844 stage (see schematics e.g. in post #7 of D3a IV stage)
I assembled it with 2 ADs stacked (output pins 6 connected through 68 Ohm resistors). But I'm not able to nullify the pin 5 offset and the voltage seems way off. I suspect my ADs are fakes. I bought them from China ebay because I have no other source for DIP package.
Anyway the DC voltages are: pin 1...-15.04, 2... 0.45, 3.... 0 (gnd), 4... -15.04 (supply), 5... -11.7, 6... -12.9, 7... 15.04 (supply), 8... 0
The JFET null circuit has little influence on pin 5, some 0.2 V, but no way to go to zero. Both channels are basically same. What do you think?
Based on your reported voltages it seems overwhelmingly likely you've been sold fakes.
The reason being the voltages you're getting on pin1 and pin8 (the offset nulling pins). My sample of AD844 shows there are ~10kohm resistors to the positive rail (pin7) at these pins yet you're seeing -15.04 and 0V there. On a real AD844 you'd be seeing ~15V on both.
The reason being the voltages you're getting on pin1 and pin8 (the offset nulling pins). My sample of AD844 shows there are ~10kohm resistors to the positive rail (pin7) at these pins yet you're seeing -15.04 and 0V there. On a real AD844 you'd be seeing ~15V on both.
Thank you, that confirms my suspicion and will help in testing future samples.
The ebay shop was adeleparts2010, but they surely proliferated to most all Chinese vendors. Will see if they at least refund money.
The ebay shop was adeleparts2010, but they surely proliferated to most all Chinese vendors. Will see if they at least refund money.
The characteristics of nulling pins are perhaps the easiest way to check for fakes - you only need an ohmmeter. I seem to recall this is a good way to check OPA627s too, a widely faked chip I believe.
PetrL,
It looks like faked to me, too. If you need AD844 in DIP and don't mind they are used. I can sell you some.
Poting
It looks like faked to me, too. If you need AD844 in DIP and don't mind they are used. I can sell you some.
Poting
Thanks for the offer, but I didn't want to take a chance any more and at once placed order at Farnell. Quite costly though. If I knew they are faked so much (*), I'd rather build the Pedja discrete stage which costs almost nothing in parts.
(*) I questioned another ebay seller, now in Europe, if he is absolutely sure he has genuine parts. He cancelled the order.
(*) I questioned another ebay seller, now in Europe, if he is absolutely sure he has genuine parts. He cancelled the order.
I can understand OPA627's being faked with say TLO71's.
But an AD844?? what has the same pin 5 TZ output that it can be faked with it??
Cheers George
But an AD844?? what has the same pin 5 TZ output that it can be faked with it??
Cheers George
I think they don't care for pin 5, the fake is not functionally equivalent to the AD844. They take whatever works in classic opamp application where pin 5 isn't used, a cheap single opamp package, maybe just the TL071.
Btw, the chinese shop refunded the money.
Btw, the chinese shop refunded the money.
It could be a TL071 - pin 1 and pin 8 show voltages that I'd expect to see for TL0x1. Pin 5 on TL071 is the second null pin and goes to VEE via ~1kohm so the 1k5 load to 0V will pull it up somewhat which is indeed what we see (-11.7V).
Pin6 at -12.7V looks to be negative saturation, looks to me to be consistent with TL0x1 but pretty much consistent with any opamp where the -ve pin is at a significantly higher voltage than the +ve. Pin5 being pulled higher than the -ve rail though might have a significant effect.
I measured the resistances of fakes which confirmed that they are not AD844:
pin 7 -- 1 ...ca. 1.5 MegaOhm (a gate-channel junction?)
pin 7 -- 8 ...open
pin 7 -- 1 ...ca. 1.5 MegaOhm (a gate-channel junction?)
pin 7 -- 8 ...open
Pin8 being open is a dead giveaway its not AD844. TL0x1 should show ~1k from pin5 and pin 1 to pin4 (VEE).
The fake shows 1k pin 4--1 (5-1 I forgott to measure).
I installed the Farnell's AD844NZ and it of course works now. TDA1541A, only 1 AD to BUF03. Very clean dynamic peaks and choruses, probably thanks to strong buffer. If anything, it lacks some bit of immediacy, but there are still several options to try. First is 3 AD stacked.
The resistances pin 7-1, 7-8 are lower than 10k, some 3-4k and different for each pin.
I installed the Farnell's AD844NZ and it of course works now. TDA1541A, only 1 AD to BUF03. Very clean dynamic peaks and choruses, probably thanks to strong buffer. If anything, it lacks some bit of immediacy, but there are still several options to try. First is 3 AD stacked.
The resistances pin 7-1, 7-8 are lower than 10k, some 3-4k and different for each pin.