Which kind of resistor?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,
I am a gain clone newbie, in the process of building my first gain clone.

This one will be a 3886 inverted design, I will make a 3875 based gain clone sometime in the future to compare.

I am hoping to build something on the higher end of the scale, and am willing to buy good parts for my gc, within reason, for example I will use Blackgate caps in it where required, probably use vampire RCA jacks and Cardas binding posts (from the group sale).

My question is on what type of resistors to use.

I have searched, and read posts (mostly by Peter) talking about Rikens and Caddocks.

I am actually thinking about buying 1Watt Rikens, Caddock TF020 and “nude” Vishays from Texas components. I would try each type in the series input and feedback resistors, and see which I like best. I guess they would all need a few days of burn in, I might do that prior to installation with a signal generator or a DC power supply.

Not the cheapest approach, will probably cost me around $70 plus shipping for these 12 resistors, and it will be a pain to try all of them out, but I don’t know of a better approach for picking out gc resistors.

Just wondering what you guys think of this, and what approaches people have used to determine what kind of resistors to use (other than trying them).

Also, have you guys played with resistor polarity? There is a thread here that talks about different resistor types, and about how the sound changed when a resistor was reversed
http://www.diyaudio.com/forums/showthread.php?threadid=6760
I would not have thought about trying to reverse a resistor before reading this thread.


BTW, this forum is very helpful for newbie GC builders.

Thanks,
Randy
 
I've been swapping Rikens to Caddocks in feedback location, so I can send you 2 Rikens 220k, 0.5W. I wouldn't recommend 1W there, as its size is too big.

You also don't have to buy all 12 resistors in the beginning. Feedback location is most sensitive to any changes, so if you like a given resistor in a feedback location, chances are it will also work well at the input.

Nude Vishays might be good, as they are small and not as damped as plastic body S102 types. I didn't like those (S102), as they sounded really mellow and were lacking body;)

Somebody also recommended tantalums from Partsconnexion, and since you are getting free Rikens, you might try those as well.;) Go for the smallest size, as this is supposed to be important.

You might try to get those nude Vishays for 10k input resistor and get only appropriate feddback types, and in case the Vishays won't work out buy same type you liked in feedback for the input. Jonathan Carr is using those nude Vishays (I presume) in his preamps volume control and they supposed to be very neutral.

I kind of stay with 10K Rikens at the input, as I already bought them and don't want to swap too many parts;)

But I will definitely be interested in your findings. As to the orientation and direction in resistors, I believe I've heard the difference, but it's quite tricky to find the right one and I'm not sure if markings or body style is consistent with the direction, so this complicates things even more. So I don't bother with that, unless the location is very special and requires no more that two pieces, like feedback or I/V stage.
 
Peter Daniel said:
I've been swapping Rikens to Caddocks in feedback location, so I can send you 2 Rikens 220k, 0.5W. I wouldn't recommend 1W there, as its size is too big.

You also don't have to buy all 12 resistors in the beginning. Feedback location is most sensitive to any changes, so if you like a given resistor in a feedback location, chances are it will also work well at the input.

Nude Vishays might be good, as they are small and not as damped as plastic body S102 types. I didn't like those (S102), as they sounded really mellow and were lacking body;)

Somebody also recommended tantalums from Partsconnexion, and since you are getting free Rikens, you might try those as well.;) Go for the smallest size, as this is supposed to be important.

You might try to get those nude Vishays for 10k input resistor and get only appropriate feddback types, and in case the Vishays won't work out buy same type you liked in feedback for the input. Jonathan Carr is using those nude Vishays (I presume) in his preamps volume control and they supposed to be very neutral.

I kind of stay with 10K Rikens at the input, as I already bought them and don't want to swap too many parts;)

But I will definitely be interested in your findings. As to the orientation and direction in resistors, I believe I've heard the difference, but it's quite tricky to find the right one and I'm not sure if markings or body style is consistent with the direction, so this complicates things even more. So I don't bother with that, unless the location is very special and requires no more that two pieces, like feedback or I/V stage.

Hi Peter,
That's very generous of you to offer the Rikens to me for testing. I'll send you some contact information.

I will go with the smaller resistors, I was going to go with larger ones because some people think that higher wattage resistors sound better, but as with everything, I am sure it's application specific.

Thanks for the advice on just changing the feedback resistors, I thought both would be important, but I guess the feedback makes more of a difference. I have to decide what to order now. I am placing most of my order with partsconnexion since they have most of what I need, so I will use the Caddock MK132's instead, since that what partsconnexion carries. Otherwise I would have to place an order with Percy just for the Caddocks. I will also try the tants, and see how they are. I may still buy the 10K inputs and 220K feedbacks in several flavors, by the time you factor in shipping/handling I am not sure if I will save much by testing the feedbacks first, and changing the signal resistor later. But just changing the feedbacks make swapping/testing easier. I could also mix and match, but that would be way too many combinations. :eek:

Oh, and don't hold your breath waiting for test results. Projects take a while for me. I have been planning my gainclone for a couple of months now, and have not started building anything yet. I have 2 small kids, so time is scarce for me. But, I finally have most of the parts, just need these resistors and a few more caps and I will have everything. I hope to start working on it soon.

Randy
 
I`ve just "finished" my first Gainclone, it`s now boxed, with separate power supply. The only thing I need is for the MUR860s to arrive so that I can build separate bridges (I`m running from just 1 transformer at the moment), I also need to order some resistors for the switched attenuator, I just use a cheap pot at the moment. The question of resistor types has been on my mind, at the moment I am using 1/8 watt metal film (1%) for the 220k feedback and an 1/8 watt carbon film 10K on the input. These parts seem to work well, my theory is that the smaller the resistor, the less resistor the signal has to pass through (probably got this wrong, but it seems to make sense !). I will also use 1/8 watt metal films for the attenuator. I am thinking of using Welwyn 1/4 watt 0.1 % resistors instead of the metal films, but they are a little more expensive (about £1 each = $1.60). Will the better quality Welwyns really make a noticeable difference or should I stick with the cheaper ones that I am using ?
 
Will the better quality Welwyns really make a noticeable difference or should I stick with the cheaper ones that I am using ?

From my experience over the last three years, I would say yes, yes, yes!

And you don't have to pay £1.00 each - just use the cheaper RC55C versions at less than half that price and believe me, you won't hear any difference. Probably the best quality resistor at the price!
 
Hi!

Nuuk said:

And you don't have to pay £1.00 each - just use the cheaper RC55C versions at less than half that price and believe me, you won't hear any difference. Probably the best quality resistor at the price!


Would I have read this earlier...

Build my GCs with the more expensive RC55 type, and just lately refitted my P3A with RC55... cost me about 25€ just for resistors for both sides...

Sadly I cannot listen to my P3A in the modified version, since one side is still broken (after I fixed it, I resoldered a broken transistor, and I think that killed the LED, and, believe it or not, although having about 70 Panasonic FCs etc. just lying around, not a single standard green LED available.... :cannotbe: :cannotbe: :cannotbe: )

If those Welwyns won't make a difference in sound, it won't matter, since they look cooler than the normal resistors I used before... :)
 
Peter:

>Jonathan Carr is using those nude Vishays (I presume) in his preamps volume control and they supposed to be very neutral.<

You mean the resistor that looks like a nude version of the S102K? My volume control is all SMD, so it isn't possible to use _any_ variant of the S102K - even the size of the wafer is too big to fit.

Nonetheless, I have tried those nude S102Ks in various prototypes, and they worked well. Like the standard S102K, noise and inductance are quite low, accuracy is very high, tempco is very stable. OTOH, I haven't used them in Connoisseur production yet - the only place that I have used any nude S102Ks (in production) is the Lyra Erodion phono stage. So far, no problems, so I may choose to use the nude S102Ks more extensively in the future.

As you probably know, my top priorities are schematic design, pcb layout and physical structures - parts are a comparatively lower priority (generally, I allow the needs of the circuit and theory to dictate the parts requirements, but once the requirements have been established, I am fanatical about using components that match those requirements as closely as possible). As a result, I never agonize over whether I should be using a Riken RMG or a Vishay S102K - at least, not in my own designs.

hth, jonathan carr
 
Would I have read this earlier...
Build my GCs with the more expensive RC55 type, and just lately refitted my P3A with RC55... cost me about 25€ just for resistors for both sides...

The number one rule in hi-fi ;) Always visit Decibel Dungeon for money saving tips!

If those Welwyns won't make a difference in sound, it won't matter, since they look cooler than the normal resistors I used before...

Sorry, I wasn't clear enough. The RC55C's don't sound any different to the more expensive RC55's but they both sound better than the standard metal film types.
 
jcarr said:
As a result, I never agonize over whether I should be using a Riken RMG or a Vishay S102K - at least, not in my own designs.


Hi Jonathan,
Your post in AA has been an inspiration to me to try and listen to different resistors brands;)
http://www.audioasylum.com/forums/hirez/messages/15679.html

As to the nude Vishays, I remember you posted some comments on them in this forum, but now I also remember that your attenuator had SMD resistors indeed. Sorry for any confusion.


For those in search of better resistors, one more link from AA:
http://www.audioasylum.com/forums/hirez/messages/93363.html
 
NOS Holco's

tiroth said:
Am I the only one using NOS Holcos? I think these will be good for my Proac Response 2.5 clones.

I still have a large pile of these left over from my glass bottle circuit projects. They are better in solid state than tube circuit because they get very noisey with time at higher voltages.
I am fairly cheap, in my Zen V4 AB 1/2 watt carbon comps were used for everything but the feedback. Here a NOS 1 watt Holco was used. An H2 I think.
The feedback position is fairly critical, but I think that resistor sound is a very small part of the overall puzzle.
To me, the use of resistors that are non-magnetic is more important than the brand or type of material.
But, I am biased, in quanities of 50 or more, the AB carbon comp 1/4 and 1/2 watt resistors are 4 cents each at my favorite surplus store. A GC requires less than 1.00 worth. I always purchase at least twice as many as needed, then match them up with a VOM. These seem to be +/- 10% even the gold band tolerance. A lot of times I will use 6 to get a matched pair.
But measuring resistors has proven to be a good thing even when using tighter tolerance resistors. I have found a small percentage of the NOS Holco that measured way off the stamped value. I used to use them in everything and never bothered checking the value. I had read them measured =/- 0.25%. They are tight but there are random flyers, either mislabeled or just culls.
I recommendation that every resistor used in the simple circuits of DIY be measured in value prior to use. This may solve some of those unexplained projects that never work just right.


George
 
Originally posted by jcarr in AA post
If anything, it is probably because they are simply looking for a different kind of sound than you are.

Jonathan,
You certainly agree, that they (Rikens, Caddocks or generic type) provide different sound in some locations. In GC we are basically dealing with one critical location and the sound changes considerably, depending on a parts choice. I'm not trying to promote one brand over the other, but simply indicate, that experimentation is worth the trouble.

Since there is not much opportunity to improve the schematic design and PCB is not required here, parts become comparatively higher priority ;)
 
Peter Daniel said:

Since there is not much opportunity to improve the schematic design and PCB is not required here, parts become comparatively higher priority ;)

Hi Peter,
I really agree with this, hence this thread.

BTW, I got some quotes for naked Vishays,

S102C/K TX2352 220K00 1.0% = $17.16ea.
0.05% = $20.34ea.
0.01% = $21.76ea.

:bigeyes: :bigeyes:

Prices were kind of a surprise to me. The 10K was $7.50, they said higher value resistors cost more. I guess so. :rolleyes:

Everyone else,
thanks for the inputs. I was not sure how much response I would get, so thanks for replying.

Randy
 
Re: and the big one?

Ricren said:
And what brand you guy use for the .22 at the output?


I don't even think you need that resistor. I never used it. Tried it once, but didn't hear any improvement, only the signature of a resistor. It was Mills 5W.

randytsuch said:


Hi Peter,
I really agree with this, hence this thread.

BTW, I got some quotes for naked Vishays,

S102C/K TX2352 220K00 1.0% = $17.16ea.
0.05% = $20.34ea.
0.01% = $21.76ea.

:bigeyes: :bigeyes:

Prices were kind of a surprise to me. The 10K was $7.50, they said higher value resistors cost more. I guess so. :rolleyes:

Everyone else,
thanks for the inputs. I was not sure how much response I would get, so thanks for replying.

Randy


With Vishays, the higher the resistance, the higher goes the price. S102 in 200k is around $30, I tried it but didn't like it. Wonder, if somehow I could take that plastic body off?;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.