Ideal I/V resistor ?

Status
Not open for further replies.
Hello,


Please: what are the items needed for a good resistor for an I/V dac chip stage ?


read many things, can't decide : Rhopoint ? Caddock ? Vishay, Allan Bradley, Sussumu smd ?


Thick film ? Naked à la Vishay or Caddock ? film type ? SMD ? carbon comp ? wirewound ? exotic ? high wattage ? Radial, axial ? Big, short ? vibration proof ? metal leads type and thickness ? Different choice according Z : 15 ohms type needs to be different than 2000 ohms for a good passive I/V


How to sort out ? Just a basic metal film ? But which ?


help & experience much appreciated :)
 
me too : JFet CCS feeded; opa861 or discrete diamonds buffer... no feedback ! My fault, should not have said "passive" as I suggested as well.


discrete serie Darlington serie with zener or very low noise ref (btw which ones today ?) for regs...


Here i'm focussed on the I/V resistor... as it's a part thread... read many things but lost in all the resistor choices (read things from Walker member,etc, asf...)


Any experience about that passive part Abrax :) ? Anyway, if a shematic I take it too, don't like the idea of an 627 with feedback for I/V than Miro chose on his AD1862 headphone dac but think it will be surely off topic in this forum area...
 
Last edited:
So far I haven't found the quality of the I/V resistor to be a significant influence on the SQ (I'm still using ultra-cheap thick film, have tried Susumu thin film). By this I don't mean to say it makes no difference, rather I mean in my experience it hasn't been the low-hanging fruit. There have been other, much bigger fish to fry - most recently that has been passive filtration of the DAC output prior to I/V.

btw 'TC' means temperature coefficient.
 
You mean filtering at the outputt of the dac before I/V stage or do you talk about filtering after for alliasing things (readed your blog :) ) ?


So Something different than the basic RC to ground for noise you meant ? Don't understand ?


thanks for the TC : ok, so less than 5 ppm...
 
Last edited:
You mean filtering at the outputt of the dac before I/V stage or do you talk about filtering after for alliasing things (readed your blog :) ) ?

I mean filtering the DAC's output before it goes into any active stage. The idea of passive I/V is that its more linear than an active stage when exposed to the wideband RF output from a DAC chip. But passive I/V doesn't give the DAC the best environment - it really needs to see a low impedance at its current output pin. I found passive I/V means the DAC's too sensitive to its power supply quality, but this could just be the DAC I'm using (TDA1387). Fact is most audio current output DACs aren't designed for passive I/V.

Yes I do mean something quite different from the basic RC to gnd. I mean something much, much better at rejecting components above 24kHz.
 
Ah, thanks ! found Nothing from you on this particular subject on current outputt filtering from a DAC.


I assume you're using smd coils, smd caps & resistor ?


It should minimize the quality of the I/V resistor I'm looking for. Basicly I readed AD844 big thread I/V (was for a PCM 1704 for I/V stage )...


Should I not talk about "passive I/V" when talking about the shunt resistor I use after OPA861 or AD844, etc ?


Well if some have experienced about this shunt resistor for I/V , I will be still pleased about inputs please.
 
Rhopoints Best I Have ever heard

I use 133 ohm Rhopoints as IV resistors for my DIY PCM1794 DAC. I also built this DAC for a friend of mine who says it sounds better than any DAC he has ever heard at any price.

Dale RNXX sounds pleasantly veiled. Susumu RG SMD and Audio Note tantalum sound clearer. The Rhopoints sound best - extremely clean, detailed, and pure sounding without being analytical or bright.
 
Ah, thanks ! found Nothing from you on this particular subject on current outputt filtering from a DAC.

If you're referring to absence of details on it on my blog, yes its quite a recent discovery I've made, since the DIYA blogs were made read-only. I have talked about it on Hackaday.io, will find you the link if you're interested.

I assume you're using smd coils, smd caps & resistor ?

The most compact, cheapest version of the filter does indeed use SMT coils - TDK SLF7045s as these are by far the best bang-for-the-buck in small inductors. There are better choices if you have more available space - P14 gapped cores - but then you'll need to wind them yourself. I use SMD caps (X7R at the lower end, NP0 for the ultimate performance) for all my filters.

Should I not talk about "passive I/V" when talking about the shunt resistor I use after OPA861 or AD844, etc ?

Best to drop the word 'passive' in this case, yes. All I/V circuits ultimately use a resistor to convert I to V. 'I/V resistor' would help minimize confusion.
 
As others have mentioned, the Susumu thin films (RG series) are very low cost yet high performance. TC is the temperature coefficient of resistance, and it's a parameter that you will pay extra to reduce.

The basic Susumu RG model is 0.1% tolerance and 25 parts per million per degree C. This is really very good performance, but you can get better performance. The benefit is that as signal passes through the resistor, the resistance film heats and cools along with the signal voltage, modulating the resistance value, and thus creating distortion. The heat cannot be changed without changing the signal levels, but if the temperature coefficient is lower, the error caused by the heat will be reduced.

This effect mainly happens at low frequencies, but it can result in measurable distortion (3rd harmonic), especially with the smaller thermal mass of a thin film SMD resistor. Choosing a 3216 size resistor will reduce this problem, as will choosing a better tempco, such as 10 ppm or 5 ppm. These days, 3216 resistors are harder to order, so buying a lower tempco with a 2012 resistor might be the most realistic approach.

This effect is why Bruce Hofer recommends against foil resistors. These resistors use a compensating thermal expansion coefficient of the resistance foil and the ceramic substrate. In other words, when the current through the resistor heats the foil, it causes it to expand, and the same heat from the foil causes the ceramic substrate to contract, mostly nulling out the effects of the heat on the foil resistance.

The problem is that the foil heats first, and then the substrate heats, so the thermal compensation has a time lag. For a DC application, there's no problem - you just wait a tiny bit for it to stabilize, and you get a really amazing temperature coefficient, far better than the foil alone. However, if you are so unlucky as to have this time lag correspond to a half cycle of an AC signal, then the compensation and the raw TC error of the foil work against each other, producing more distortion than if you simply used a better resistance film. This effect is purely at low frequencies, practically around 50 to 200Hz as a rough guess. With higher frequency signals, the heating and cooling take place too quickly to modulate the foil resistance, minimizing this effect.

As was also stated, there are bigger fish to fry, and as long as you take the thermal issues into account, you don't have to spend a fortune on a resistor to have it work really well.

Filtration of the DAC output prior to the I/V stage is very useful, as is providing a virtual earth load to the DAC output. A simple resistor I/V stage will not properly 'pamper' the DAC stage with a nice VE load, and will probably cause the DAC to misbehave more than if you simply used a quality virtual earth I/V amplifier stage.

There are modern low voltage op amps that can handle getting blasted with tons of HF trash from the DAC, and can provide enormous gain-bandwidth, in order to create a wideband, low distortion virtual earth. My pick would be to use one of the modern several GHz gan-bandwidth low voltage op amps on a really nice multilayer PCB with some cleverly done power supply bypassing and passive lowpass filtering of the DAC output. I personally think that much of the magic is not the simple I/V resistor, but instead in pampering the DAC output and extracting the signal carefully with some well designed passives and a well layed out high speed I/V amplifier.
 
Very instructiv, many thanks.


Some transconductance à la opa861 - so 5V low voltage- (so without feedback for speed instead the opa860?) to allow the dac current output to see a low Z is what you refer to ? I use it without knowing if better these days on manufacters shelves ?


Or is there wider bandwith stuffs : aop for video ? or simpliest transistor ?


Any ref in mind (it is still a parts thread :) )




ahaha, wanted one week ago to buy this instead a naked Vishay : Mouser Electronics, Inc. Europe==
 
Last edited:
Status
Not open for further replies.