• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Can I increase capacitance on these TONE BOARD caps?

Mr Kantor, if you are curious there is a link to a full schematic on post #19 of this thread. The amp is a Kenwood KA-4006 integrated circa 1974. There is a similar 4558 opamp on the phono pre-amp. It is my understanding that the opamp under discussion in this thread, on the tone board, is in the signal path at all times. That is why it has my interest.

There are no other ICs in this amp.

I would also note that the main amp board does have a speaker protection relay. Your comments about turn on/off thumps make me wonder why the opamp circuit would perform this function when, to my understanding, the speaker relay handles this. But I am here to learn.
 
This is also the way I see the situation.

I think Kenwood put additional filtering on the tone board to improve the regulation from the zeners. For this, a series resistor (R127/8) was needed with the consequence that the supply voltage on the tone board is reduced to something near ±9V.

The tone board is after the volume control, and just before the main amp. The supply voltage on the main amp is less than ±50V, and the voltage gain of the main amp is close to 11. So, the power amp will clip with an input signal of 4.5V peak. I don't see any reason to increase the supply voltage on the tone board. You already have 6dB headroom on the tone board with the actual ±9V supply.
 
roger2 said:
...Having found that TO-99 op amps are still available (linked in my last post) I would prefer to explore that option first. I would need guidance on how to choose from among those op amps available...
Click to expand...

Surprisingly, there were TO-99 options available, and even one that EchoWars could endorse. Unfortunately my lack of oscilloscope availability and operating knowledge requires that I step back and re-evaluate.


@ EchoWars

premise: if I am able to solve the mounting issues and any standard DIP-8 opamp becomes an option

I understand that any opamp swap is a non-trivial undertaking. And I am aware that guys "roll" opamps all the time, never scope them, and then have issues that they will never even be aware of in many cases. I do not want to do that.

But I am wondering if the complexities involved grow greater as the difference in the opamps in question becomes greater?

So here is my question: What if instead of using an opamp several generations advanced from the 4558, I were to use a device from a generation nearer to the 4558. My main goal here is simply to get rid of a noisy opamp. Surely some of the successors to the 4558 solved this issue?

Again, I am fishing here...but if I swapped in a device with parameters more similar to the 4558 would it become less likely that oscillation or other issues would develop? Assuming that decoupling as mentioned previously was applied, is there such a device that might be used safely without the requirement of being scoped?

----------------------------------------

I have followed up a bit on one of nashvillebill's ideas. Got a RadioShack 741 DIP8 and a pair of IC sockets. Bent the legs of the 741 one of the sockets inwards to roughly match the holes in the PCB. Held each device next to the edge of one of the boards. It appears that the legs of the socket, bent this way, might barely protrude through to the back side but not enough to solder.

The legs on the 741 are longer. They appear to be long enough to solder to the board. But they are not really long enough to afford a good anchor for any decoupling caps. Plus I would be very concerned about heat if I tried a combination joint like that. This PCB has pretty flimsy traces.

This approach might work if there was DIP-8 socket with longer legs. Does anyone know if DIP-8 sockets exist that have legs longer than the Radio Shack units?
 
Last edited:
BinaryMike said:
Odd that you should mention the 318, Ken. I've used it in audio circuits as well, because I was very familiar with it from my work in video. I've also used video buffer chips like the LM6321 to upgrade older preamps (as main output and tape output buffers).
Click to expand...

Cool! What kind of video stuff did/do you design? The 318 made a very good audio amp... much faster and quieter than its peers at the time. Low THD, etc. It just took a little work to get it stable. If I recall, it often took a cap from V+ to V- in addition to the usual bypassing, as well as >220 Ohms on the output, maintaining a gain >3, etc. Is your experience similar?

When I was an undergrad, I didn't have access to equipment capable of measuring the THD of the 318-based designs I was working on. So, I would wait with anticipation for the traveling McIntosh clinic to roll through town. The Mac engineers kept checking their test gear, scratching their heads, and then asked if they could see what I had built my amp with. Finally, after talking circuits and amps with them for a while, and learning a bunch, I opened the box up for them to look at. I got the sense that they had never seen, or measured, an high performance IC-based amp up to that time. Phono stage used a low noise FET in front of the opamp, though.

-k
 
roger2 said:
Mr Kantor, if you are curious there is a link to a full schematic on post #19 of this thread. The amp is a Kenwood KA-4006 integrated circa 1974. There is a similar 4558 opamp on the phono pre-amp. It is my understanding that the opamp under discussion in this thread, on the tone board, is in the signal path at all times. That is why it has my interest.

There are no other ICs in this amp.

I would also note that the main amp board does have a speaker protection relay. Your comments about turn on/off thumps make me wonder why the opamp circuit would perform this function when, to my understanding, the speaker relay handles this. But I am here to learn.
Click to expand...

Thanks, I missed that!

1- This seals it for me. If the RC under discussion was for supply filtering, one would be needed on the 4558 in the phono section, which is MUCH more sensitive to supply ripple than the tone control is.

2- I only ran through the circuit quickly, so I don't guarantee this, but: oddly, it seems like the relay is Normally Closed, and opens if there is a DC or other error present. It might be closed at turn on. It's a bit hard to tell the way the circuit is labeled, so I might be wrong here.

3- Even with a traditional NO relay, you have a problem at turn off. The relay is slower than the circuit, so you need to give the relay time to open before pops are generated.

4- You could put an NJM4559 in for some improvement in the audio performance, and almost no risk. A 4560 would be a bigger step, but you would be running some risk of oscillation requiring a good scope to track down. A 5532 might work, but could well require adjusting Ri27 and Ri28. This could be done with DVM, no scope. Moving up to boutique chips from National, TI and Analog Devices requires, I think, a good scope and knowledge of the subtleties of finding and taming instabilities.

5- I think it would be a good idea to do the cap first, understand what is going on, and tackle the opamp second. Of course, everyone who rolls opamps has nothing but praise for each step, but I'm a not sure you will get any sonic revelation from this approach. The whole idea of opamps is to take the chip's inherent performance out of the picture. This is done by providing massive gain, most of which can be "thrown away" in order to achieve an almost ideal transfer function.

6- The key to getting DIP and miniDIP sockets with very long legs is to find "wire wrap" sockets.

Have fun!

-k
 
Last edited:
ken kantor said:
1- This seals it for me. If the RC under discussion was for supply filtering, one would be needed on the 4558 in the phono section, which is MUCH more sensitive to supply ripple than the tone control is.
Click to expand...
There is... Cd16 & 17.
2- I only ran through the circuit quickly, so I don't guarantee this, but: oddly, it seems like the relay is Normally Closed, and opens if there is a DC or other error present. It might be closed at turn on. It's a bit hard to tell the way the circuit is labeled, so I might be wrong here.
Click to expand...
You are...the contacts used are the NO contacts (the relay has both NO and NC contacts...on many of these old units, the mfgrs bought relays that were made with the NC contacts absent and the PC board may not even have holes in it for the NC contacts. With a new relay, the NC solder pins are simply clipped off to fit it on the board). Relay opens if DC is detected.
3- Even with a traditional NO relay, you have a problem at turn off. The relay is slower than the circuit, so you need to give the relay time to open before pops are generated.
Click to expand...
I'm not sure I got what you were trying to say, but of course, what we want to happen is for the relay to open as soon as possible once power is shut down. Most protection circuits have a simple AC sense circuit tied to the base of the relay driver transistor to pull it to ground once the loss of AC power is detected. This protection circuit has nothing like that, and simply relies on Ce28 discharging quick enough to open the relay once power is cut. For this reason, it is probably a good idea to stick with the stock values for Ci15 & 16, and Cd15 & 16.
4- You could put an NJM4559 in for some improvement in the audio performance, and almost no risk. A 4560 would be a bigger step, but you would be running some risk of oscillation requiring a good scope to track down. A 5532 might work, but could well require adjusting Ri27 and Ri28. This could be done with DVM, no scope. Moving up to boutique chips from National, TI and Analog Devices requires, I think, a good scope and knowledge of the subtleties of finding and taming instabilities.
Click to expand...
True enough...without a scope you're flying blind. I can think of a LOT of fairly fast (10 to 20V/µS) that run peachy with even very poor bypassing, but there's never a guarantee. The safe suggestion is to keep the stock opamp in place, as you (roger2) simply don't have the tools needed to be tackling anything very deep here. And most opamp swaps will require that Ri27 & 28 be changed IF you wanted to keep the stock supply voltage to the chip.
6- The key to getting DIP and miniDIP sockets with very long legs is to find "wire wrap" sockets.
Click to expand...
Yep...but I can't remember how stiff and/or 'easy to bend' the leads of a WR sockets are...haven't seen one in ages. They are still sold, but I simply haven't had any reason to use one in the last 20 years. I am surprised that the PC board holes are so 'narrow'. Hmmmm.
 
EW- "it is probably a good idea to stick with the stock values for Ci15 & 16" is all I have been trying to say from the very beginning!!!! Perhaps we agree on this?

Other comments:

EchoWars said:
There is... Cd16 & 17.
Click to expand...

Really?? That network has only 1/6th the impedance, even on a more sensitive section. Corner is much higher. Only 1V is dropped on the V+ and V- lines.

So what is the use of 680R/100u on the tone section, which is run off the same supply?? If there is supply ripple getting into the signal 20x more of it is getting in at the preamp! Why didn't they use 680R/100u there?? Plus, voltage swing is not needed nearly as much as the tone section.

I'll try and mock this up on a protoboard over the next week, and run some measurements to post.

EchoWars said:
You are...the contacts used are the NO
Click to expand...

OK, it happens... can you explain the relay control circuit; especially the portion near the diode bridge? Thanks!

-k
 
Last edited:
ken kantor said:
Cool! What kind of video stuff did/do you design? The 318 made a very good audio amp... much faster and quieter than its peers at the time. Low THD, etc. It just took a little work to get it stable. If I recall, it often took a cap from V+ to V- in addition to the usual bypassing, as well as >220 Ohms on the output, maintaining a gain >3, etc. Is your experience similar?
Click to expand...

Most of my product design work has been in that middle ground between pro/broadcast and consumer video -- signal processing boxes, special effects generators, and lately, character/graphics genlock-overlay modules. The 318 was indeed a bit ticklish to work with, but I seem to recall using them successfully at unity gain once or twice. They run out of bandwidth pretty quickly when you want a substantial amount of video gain. Most video op-amps need excellent power supply decoupling. With 318s, I used tantalums and ceramics in parallel on both rails. My newer designs often run on a single +5V rail, which simplifies power supply arrangements but complicates dynamic range issues. The OPA354 is my favorite analog video go-to chip these days. None of this is relevant to the OP's questions, so I think I'll try to focus on 'business' now...
 
ken kantor said:
EW- "it is probably a good idea to stick with the stock values for Ci15 & 16" is all I have been trying to say from the very beginning!!!! Perhaps we agree on this?
Click to expand...
After looking at the protection circuit (which is awfully simplistic, whichever version the amp actually has), I'd have to agree that staying with the stock values is probably a good idea to make sure that the unit behaves properly at power-off. Like I said to begin with, there really isn't much to gain by increasing the value anyway.
 
EchoWars said:
After looking at the protection circuit (which is awfully simplistic, whichever version the amp actually has), I'd have to agree that staying with the stock values is probably a good idea to make sure that the unit behaves properly at power-off. Like I said to begin with, there really isn't much to gain by increasing the value anyway.
Click to expand...

I just saw it. One of these days, I will build and test the circuit(s) in question. I wonder if I can even measure power supply influences on the National chip. I'll keep you posted!

-k
 
BinaryMike said:
Most of my product design work has been in that middle ground between pro/broadcast and consumer video -- signal processing boxes, special effects generators, and lately, character/graphics genlock-overlay modules. The 318 was indeed a bit ticklish to work with, but I seem to recall using them successfully at unity gain once or twice. They run out of bandwidth pretty quickly when you want a substantial amount of video gain. Most video op-amps need excellent power supply decoupling. With 318s, I used tantalums and ceramics in parallel on both rails. My newer designs often run on a single +5V rail, which simplifies power supply arrangements but complicates dynamic range issues. The OPA354 is my favorite analog video go-to chip these days. None of this is relevant to the OP's questions, so I think I'll try to focus on 'business' now...
Click to expand...

I've done very few video design jobs, so am hardly knowledgeable in the area. But, I do remember often having to use discrete circuits in order to get the bandwidth required. Chips like the 354, the 6714 and some of the current-feedback stuff are just remarkable to me.

-k
 
In this particular situation it would be easy to built a small adaptor from a piece of perforated board and a few inches of solid copper wire (telephone). We just need to adapt a 0.3 inch X 0.3 inch socket to the 0.3 inch X 0.2 inch holes.
 
Based on Ken Kantor's suggestion I Googled wire wrap DIP sockets and got several hits. I have not even investigated them all yet. But the second item on this page looks like what I might need.



ecluser your idea seems doable too. Do you mean that short wires would be soldered to the pins of a standard socket?


-------------------------------------------------
Regarding a possible replacement for my 4558: I am digging through datasheets for all available 4559 and 4558 DIP-8's (Mouser and Digikey) and have been surprised at how much variation there is in the specs for the same IC but from different manufacturers. One 4558 has a noise spec of 8nV which seems not too far worse than the 5532 spec (5nV) BinaryMike mentioned.

I am also encountering difficulties because not all datasheets have every spec, and some of the specs are specified using different units, noise in particular. Once I get a decent summary of specs I will compile it in a spreadsheet and post here including spec sheets.

One question about this search though: TI has both 4558 and 4559 series using a package called PDIP instead of DIP. The physical dimensions look the same as a normal DIP. Also, Mouser's description field refers to these as "Hi Performance". Are these TI PDIP opamps different in some way that would eliminate them from consideration?
 
Last edited:
roger2 said:
ecluser your idea seems doable too. Do you mean that short wires would be soldered to the pins of a standard socket?
Click to expand...

I would solder the standard socket on a perforated board with pin extensions (single strand wire inserted in the same hole as the pins of the socket) on one row, and solder 4 extensions on a row adjacent to the second row of the standard socket. Does my approximate english make sense to you?
 
um...I have had a very long day. I don't understand what perforated board is in this context. Is this generic PCB material?
 
roger2 said:
One question about this search though: TI has both 4558 and 4559 series using a package called PDIP instead of DIP. The physical dimensions look the same as a normal DIP. Also, Mouser's description field refers to these as "Hi Performance". Are these TI PDIP opamps different in some way that would eliminate them from consideration?
Click to expand...

PDIP is just a plastic (epoxy) DIP, which is the industry standard package. Ceramic DIPs are sometimes available for aerospace and military applications.

I think that "High Performance" moniker was gratuitously applied by Mouser, even though TI's RC4558 types appear on a Mouser catalog page titled "Standard Linear Amplifiers" and we know that RC4558 was originally just Raytheon's modestly improved version of the ancient general-purpose 741.
 
You must log in or register to reply here.
Back
Top Bottom