UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 17th Jun 2019, 11:55 am   #41
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

I have no personal interest in Khozmo, Joe - just providing the link. So please don't bust my balls because you use a different switched attenuator.

Craig
Craig Sawyers is online now  
Old 17th Jun 2019, 12:24 pm   #42
Electronpusher0
Nonode
 
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,287
Default Re: Valve preamp.

Quote:
Originally Posted by Electronpusher0 View Post
Hi Andy, you have got me curious, I will put the pre-amp back on the bench and check the output again.
In my setup the volume control is at the input of the main amp, so after the cathode follower in the pre-amp.

Hi Andy,
I put the pre-amp on the bench as promised. I attached a modified schematic, seems I used a zener stabilised HT supply when I built it. I also used a regulated DC supply for the heaters and even fitted mesh screening over the mains lead! Photos attached.

As to results. I deliberately kept the gain down by using an ECC82 in the V2 position, this yields an overall gain of 65 at 1KHz ie a 5mV input will give 325mV output which is perfect for my application. However clipping occurs at 29V rms output!!!!! yes thats volts.

I tried changing the ECC82 for an ECC83 to see it would be more suitable for your application.
This changed the Anode voltage from 91V to 205V not surprisingly.
The gain of this version is now 291 at 1KHz which would give 1.5V for 5mV input, much closer to what you need.
Clipping now occurs at only (only!) 21V rms which represents an input of 72mV rms, which is 23dB above the 5mV nominal


Peter
Attached Thumbnails
Click image for larger version

Name:	IMG_1875 (Large).jpg
Views:	156
Size:	100.5 KB
ID:	185245   Click image for larger version

Name:	IMG_1876 (Large).jpg
Views:	177
Size:	118.6 KB
ID:	185246  
Attached Files
File Type: pdf Valve RIAA preamp modified.pdf (161.4 KB, 129 views)

Last edited by Electronpusher0; 17th Jun 2019 at 12:32 pm. Reason: spelling
Electronpusher0 is offline  
Old 18th Jun 2019, 1:33 am   #43
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Valve preamp.

I must say Peter, I DO like your design very much. It's simple, elegant and has no audiophool frills. I am very tempted to give it a try myself if you don't mind me plaigerising your design. To Craig, my comments were tongue in cheek, as it reeks of the audiophool thread. I did not mean any slight!!.

Joe
joebog1 is offline  
Old 18th Jun 2019, 7:01 am   #44
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Valve preamp.

Thanks lads and thanks again Peter for going out your way to help. I see you paralleled the caps in your RIAA filter, I was wondering how to get 8n2 and 560p, though thinking about it am sure I have some 560p silver mica caps.

I built Fred Nachbaur's phono amp in the interest of science and all that, attached schematic below again, but found it to lively. Just tweaking it to get the DC conditions right took a lot of doing and I still couldn't get it bang on. Also the OP was "boingy", a bit like a wein bridge oscillator with a lamp as the gain control. So I won't be building it.

The one thing that has taken a lot of time on this project is finding caps, I have draws full of the ******* but still had to root through the sheds and parallel caps to get odd values in schematics.

This order of circuits has been suggested - RIAA ( with it's own buffer ) , input selection, attenuator ( volume control), line amp + tone controls, output buffer. Another good idea is to put pads or trimpots on IP's, 47k is the usual line level IP Z isn't it? I have some 100k and 50k pots that should do the job there.

Andy.
Attached Thumbnails
Click image for larger version

Name:	Fred Nachbauer's phono stage.jpg
Views:	114
Size:	99.0 KB
ID:	185279  
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 18th Jun 2019, 8:02 am   #45
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

I've just run a passive RIAA through TINA, allowing 0.1% tolerance to perfect value resistors, and 1% to capacitors.

The maximum error, with all tolerances stacking up worst case, is +/-0.094dB.

Picking the nearest E192 value resistors in 0.1%, and two standard value caps in parallel for one of the caps (all 1%), gives +/-0.11dB worst case stack up, which is perfectly acceptible.

That is for the passive RIAA that I use, so series R of 22k, then 3.2k in series with 100n, then 33n//1n across that.

Craig
Attached Thumbnails
Click image for larger version

Name:	RIAA tolerance.jpg
Views:	84
Size:	76.6 KB
ID:	185280  
Craig Sawyers is online now  
Old 18th Jun 2019, 8:30 am   #46
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

Quote:
Originally Posted by Electronpusher0 View Post
Hi Andy,
I put the pre-amp on the bench as promised. I attached a modified schematic,
Just put the RIAA values (without tolerances) from the schematic into TINA. The response deviates quite a lot (+/-0.8dB), with a big hump at low frequency. From 100Hz to 20kHz it is +/-0.15dB.

However, keeping the 330k fixed, take out the 560p and just use 2.2n. Change the 39k to 48.1k (E192 standard value) and the cap in series with that becomes 5.6n//1n (for 6.6n).

That gives +/-0.13dB from 20Hz to 20kHz assuming perfect values

Craig
Craig Sawyers is online now  
Old 18th Jun 2019, 8:50 am   #47
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Valve preamp.

I must really retire!! When I was a kid (at uni) 1dB was the smallest detectable change that the human ear could perceive, and that was for volume level. Also that was under ideal laboratory conditions. What about the "tone controls" set by the recording engineer??

I will put a phew dollars on it, that the guy running the mixer, mixed what he considered a good sound!! Be damned with RIAA!! BUT when it's mixed down and "compressed" to vinyl format, does it really matter? .13 dB deviation from the ideal ( and I MUST add, in YOUR laboratory only) coupled to a ball raced volume control, must be better than the London Symphony Orchestra, live!!
Within an RIAA curve .

My last post was deleted as being rubbish (as are 99% of my posts) But is .13dB deviation anything to even consider? With my respects!!.

( A local electronics magazine called Radio, Television and Hobbies, once had a regular monthly discussion, called "Lets Buy an Argument" Perhaps the convenors of this site would consider an equivalent?

Joe
joebog1 is offline  
Old 18th Jun 2019, 9:50 am   #48
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

You do have a somewhat antipodean bluntness Joe. Or maybe you just don't like me.

The question is - since it is so easy do RIAA correctly with stock components that cost next to nothing, why on earth wouldn't you?

Craig
Craig Sawyers is online now  
Old 18th Jun 2019, 11:02 am   #49
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,798
Default Re: Valve preamp.

Quote:
Originally Posted by Diabolical Artificer View Post
I built Fred Nachbaur's phono amp in the interest of science and all that, attached schematic below again, but found it to lively. Just tweaking it to get the DC conditions right took a lot of doing and I still couldn't get it bang on. Also the OP was "boingy", a bit like a wein bridge oscillator with a lamp as the gain control. So I won't be building it.
It's only a phono preamp. No rocket scientists need to be harmed in the creation of one. Trouble just setting the DC conditions does not give any confidence in the quality of a design. And the 'boingy' description, while it is both poetic and liable to attack as undefined, sounds to be a strong indication that the circuit is of only marginal stability. File under a-for Avoid. With a capital 'A'!

We are in the 21st century and it's important to consider things in context. THere are NE5534 phono stages which offer low noise, low cost and just work. They've been on the go long enough to be considered certainly classic if not quite vintage yet.

So given the existence of the NE5534 and things like Hugh Walker's 3-transistor job, why use valves?

1) If you want to learn about valves. Fine! Good! Great! But maybe you want to learn from a good circuit and not something with major problems? However you'd learn a lot from analysing those problems but this would be entirely a learning exercise and would leave you with something you would not want to use.

2) Just for the hell of it. Fine, Good fun! It can make something which can be a conversation piece and mystify visitors.

3) Because careful analysis and measurement show that valves can perform better than semiconductors in this application.... No, sorry, not generally true. Some good valve designs are better than some bad semiconductor designs but that gives no guide to the value of good semiconductor designs. Building a bad valve design is pointless unless you are studying the design errors.

4) Because you believe valves have stange and wonderful properties that science just can't explain and these translate into a wonderful sound which only true aficionados have the skilled hearing to be able to detect. Yes, well let them get on with it. Experience shows that such aficionados can never be convinced otherwise. They are best left to enjoy the specialist websites and magazines which cater to their beliefs.


There are some god semiconductor-based designs that are well-proven and sure-fire. You'll have to work quite hard to do a valve design which can equal them, but you could get to the point where no-one with normal hearing should be able to distinguish between them.

I agree with Joe that when you get to sub-dB differences it's time to stop worrying. THe recording studio probably had a guy setting powerful equalisers to wherever he liked the sound - entirely by ear. The record-cutting department would likely use a precision network to do the cutter equalisation, but the combination with the settings on the desk make the overall shaping less certain.

I agree with Craig that when analysing things, you might as well do the arithmetic with enough precision to be far better than what can be done in reality. If you don't, you add errors which then become significant.

My hearing has limitations. This gives me freedom to engineer within those limitations. But sometimes I like to engineer things to stupid levels of performance, just for the sheer hell of it, just to prove that I can. It's the same freedom that lets people choose what technologies they wish to use.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 18th Jun 2019, 6:38 pm   #50
Electronpusher0
Nonode
 
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,287
Default Re: Valve preamp.

Quote:
Originally Posted by Craig Sawyers View Post
Just put the RIAA values (without tolerances) from the schematic into TINA. The response deviates quite a lot (+/-0.8dB), with a big hump at low frequency. From 100Hz to 20kHz it is +/-0.15dB.

However, keeping the 330k fixed, take out the 560p and just use 2.2n. Change the 39k to 48.1k (E192 standard value) and the cap in series with that becomes 5.6n//1n (for 6.6n).

That gives +/-0.13dB from 20Hz to 20kHz assuming perfect values
Hi Craig, you really need to model the whole amp not just the RIAA components (I did this in LTspice)
The following is from my design notes:
"The (passive RIAA) circuit is affected by both the output impedance of the preceding stage and the load impedance of the following stage.
• The output resistor of the preceding stage effectively adds to R1’s value
• The input impedance of the following stage is effectively in parallel with R1
Overcoming this problem requires selecting R1’s value such that the combined effects of the preceding stage output impedance appearing in series together with the next stage input impedance appearing in parallel return it to the designed value.
Fortunately, valve amplifiers circuits have significant output impedance. For example, a 12AX7 used in a grounded-cathode configuration with a 150k plate-load resistor results in an output impedance of roughly 44k (62k in parallel with 150k)."
The R1 referred to is the input resistor of the passive RIAA.

I attach an LTspice plot of the whole amp using a reverse RIAA input.

The real world test on the bench confirms this.

Peter
Attached Thumbnails
Click image for larger version

Name:	RIAA response.jpg
Views:	77
Size:	70.6 KB
ID:	185321  

Last edited by Electronpusher0; 18th Jun 2019 at 6:42 pm. Reason: added note
Electronpusher0 is offline  
Old 18th Jun 2019, 10:52 pm   #51
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

Good point Peter. I'm more familiar with either op-amp or discrete semiconductor circuits, where you can cheerfully ignore output impedance. An output impedance of 44k is three orders of magnitude (or more) than I am used to.

Anyway, good to know that the whole circuit models just fine.

Craig
Craig Sawyers is online now  
Old 19th Jun 2019, 12:53 am   #52
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Valve preamp.

I apologise for my bluntness Craig, Its just the way I express myself I guess.
I do engineer to my best standard but Im still going to be limited by the components I can source. The best capacitors I can get hold of easily are usually 1%, resistors I can get to .1%. Then when I change a valve for whatever reason it will be slightly different in performance and that will again change the error of the RIAA filter.

Joe
joebog1 is offline  
Old 19th Jun 2019, 6:49 am   #53
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Valve preamp.

"Bang on" is not really the right term, I'm familiar with valve circuits not having a close tolerance but the Vgk measurements were well off putting the ECC83 near cutoff , couldn't get the first stage diff amp to balance too.

Am using valves David for some of the reasons you mentioned (certainly not because I think they're magic) but mainly cos I've got the bits but mostly because I've had the mains tfmr's for a while knocking about crying out to be used. I've not written off the idea of using opamp or fet based design for the phono stage, just experimenting at present.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 19th Jun 2019, 7:21 am   #54
Electronpusher0
Nonode
 
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,287
Default Re: Valve preamp.

Quote:
Originally Posted by Diabolical Artificer View Post
"Bang on" is not really the right term, I'm familiar with valve circuits not having a close tolerance but the Vgk measurements were well off putting the ECC83 near cutoff , couldn't get the first stage diff amp to balance too.

Andy.
Hi Andy, just to be clear, I assume you are talking about the circuit you posted in post #44

Peter
Electronpusher0 is offline  
Old 20th Jun 2019, 6:41 am   #55
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Valve preamp.

Yes Peter, I wanted to try an active phone stage. So far your circuit and the valve wizard one both work well, arn't too complex and used common parts. After a bit of searching I found an RCA active circuit that I'll give a try before settling on a final circuit.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is online now  
Old 20th Jun 2019, 7:07 am   #56
Electronpusher0
Nonode
 
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,287
Default Re: Valve preamp.

OK Andy,
I took a look at F. Nachbaur's pre-amp and read his write up on it.
I think one of the problems with the design, which he refers to, is that the differential amp front end has a poor common mode rejection due the fact that the common cathode load is not close enough to a constant current. He offsets this shortcoming to some extent by unballancing the Anode resistors.
Differential amps of this type are often called "long tailed pairs" because to get a close approximation to a constant current you really need a negative supply, say 120V and a larger cathode resistor, hence the long tail.
By only using a positive supply and returning the cathode resistor to 0V the design has to have the grids at about 27V which adds complexity with dc blocking capacitors.
This combined with direct coupling to the next stage means that the HT supply has to be so high, 430V!
With a negative supply the HT could come down to about 340V.

Another advantage is that the grids of the diff amp can now be at 0V which eliminates some components, the bottom of R102 can go to 0V, C102 goes as does R103 and R104 becomes a link.
There will be some changes in value eg R105 (the cathode resistor) increases to 120K.
There will also need to be a few mods to the dc feedback circuit because of the change in grid voltages, R 110 becomes 270K and goes to the -120V instead ov 0V.
The reduced HT means one of the neons is not needed, or repace the neons by a single 120V zener.

I will try and find time to model the version described above in LtSpice (I may even knock one up out of curiosity).

Peter
Electronpusher0 is offline  
Old 20th Jun 2019, 1:50 pm   #57
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,798
Default Re: Valve preamp.

Ah, it would all have been so much easier if only someone had invented a complementary valve. Asimov predicted positronics, but it has failed to manifest itself.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 20th Jun 2019, 4:19 pm   #58
Craig Sawyers
Dekatron
 
Craig Sawyers's Avatar
 
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,939
Default Re: Valve preamp.

Audio Research developed a range of topologies for RIAA back in the 80's to 90's. These usually had a singe JFET input, then some combination of triodes (sometimes just a single one), a complementary cascode MOSFET stage, then a MOSFET buffer. The RIAA was wrapped around that lot.

They clearly had either power on latch up, or latch up on overload, because there are multiple back to back diodes and zeners, and JFETS configured as diodes (ie D&S connected together) all over the shop.

The schematics can be found on line for anyone remotely interested.

Not suggesting for one minute that this is the way to go, just saying that is what one manufacturer did.

Craig
Craig Sawyers is online now  
Old 20th Jun 2019, 5:16 pm   #59
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Valve preamp.

Hi Andy,

....neat work, as usual!

What do you call the type of tagstrips you are using there? I like them a lot!
__________________
Al
Al (astral highway) is offline  
Old 21st Jun 2019, 6:46 am   #60
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,633
Default Re: Valve preamp.

Thanks Peter, I did think of using a CCS and as you said he mentions some of this in his blurb, even with the changes suggested I wouldn't use it because of relative complexity and component count.

I tried this circuit yesterday - see attached - this is the sort of circuit I like, no weird odd value caps, small component count etc. Anyway it worked ok on the 1 x 10 and 1 x 100 range of my sig gen with a quick test, but has a lot of top end. At 10khz amplitude is four times as big as 1khz, I will test it more thoroughly today and change some values on the FB path.

Regarding using FET's, I got a few of the 2SK369 low noise jobbies to try Pat Turners Rocket circuit ( without the paralleled caps!) but after reflection knocked it on the head - too many valves & overly complex. PT built his Rocket amp after reading Secrets of the phono amp stage by Allen Wright - attached. Audiof*ckwitery? It makes sense to use a low noise FET I guess with a following valve stage to get The Tone but most 2SK369/2SK170 phono amps I've looked at lean towards Audiow**kery and treat the FETS as sacred.

Thanks Al, wondered where you got to and how you were. They're ceramic wafers out of a scrap Tektronix 545B scope, used a lot in all the 500 series scopes.
Lasly there's lots of useful circuits in this - http://www.tubebooks.org/tubedata/RC30.pdf

Andy.
Attached Thumbnails
Click image for larger version

Name:	shure m65.jpg
Views:	440
Size:	93.4 KB
ID:	185441  
Attached Files
File Type: pdf SP_15_Article.pdf (661.6 KB, 106 views)
__________________
Curiosity hasn't killed this cat...so far.

Last edited by Diabolical Artificer; 21st Jun 2019 at 7:06 am.
Diabolical Artificer is online now  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 10:01 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.