28 replies to this topic

[Pete B]

I was given an Advent 300 Receiver many years ago by a coworker who said it was broken and not worth fixing. I offered to fix it but he wasn't interested. He said it thumped loudly on power up/down and something was strange with the tuner. I did nothing to it but used it on occasion as a spare amp.
I was never much interested in this amp as it does not have enough power for a main system, and the build quality is rather low.

I'm rebuilding it now just to have a working sample of the phono section, and will fix the line section so that it can be used as a full preamp.
No plans to work on the tuner or power amp at this time.

Verified operation, amp output offset low.
Power supply voltage good.
Unit pops badly on power up/down.
Some controls are noisy.

1. Sprayed all contacts with Deoxit
2. Replaced pair of 3000 uF (date code 22 week of 1977) power supply caps with a pair of 10,000 uF.
3. Replaced all line section PSU electrolytic caps with tantalums, I do not normally suggest these but I had some good quality hermetically sealed units.
4. Replaced all Line signal path caps with GE poly types.
5. There is no tone control defeat, hard wired tone controls out of circuit.
6. There is a FET based mute circuit on the output of the line preamp to avoid power up thumps and pops. This was not working because the cheap electrolytic timing capacitor was far out of spec, if not completely shot. Replaced with a tantalum and at least the most offending pops are gone.

Some comments based on the schematics:
1. Interesting that the line stage is inverting.
2. Volume pot is 25K ohms, which is low and approximately defines the line level input impedance.
3. Volume pot drives the tone control section. The line level input impedance is flat with tone controls set flat, but becomes frequency dependent with non-flat tone settings. I verified this through SPICE simulations.
4. Pre amp power supplies employ active regulation, line +/-15V, phono +/-12V.
5. Line stage is a primitive transistor based OP amp with a simple diff amp and VAS employing 3 transistors.
6. There is no RF filtering, line stage is -3dB at 1Hz and 490 kHz as predicted through simulation.
7. Phono section utilized Holman's high pass filter at output of preamp rather than the input as described in the AES article.
8. Split supplies are used with OP amp type amplifiers which could offer direct coupled signal paths if attention was paid to DC offsets, however capacitor coupling was used instead.
9. PC board layout is haphazard making it difficult to locate parts.

Pete B.

[Pete B]

I came across this web page covers a rebuild of the Advent Model 300 receiver:
http://home.netcarri...ens/advent.html

Schematics here:
http://www.davidreat...r_main_page.htm

Tomlinson Holman's AES paper covering the phono section in .pdf format:
http://www.davidreat...n_AES_paper.pdf

Audio Critic review in .pdf format:
http://www.davidreat...eamp_Survey.pdf

Specs:
http://www.davidreat...cifications.pdf

I spent some time looking at the schematics years ago, and am now taking another look.

Pete B.

[Pete B]

I mentioned that differential amplifiers work better, have lower distortion, when used in the inverting configuration as was done in the line stage. This is because the positive input is grounded, thus the negative input becomes a virtual ground, and therefore the diff amp base connections do not swing any large voltage that would otherwise modulate the device capacitances causing increased distortion. It acts in a sense as a current amplifier. I wrote more about this configuration here:
http://www.diyaudio....?threadid=48873

Some worry about absolute phase and might see the Advent line stage as having a disadvantage in this regard but the simple solution is to reverse the phase of both speakers. This does not correct for headphone listening which is another consideration.
It is odd that they did not make the power amp inverting to have the two sections compensate for one another and to use a consistent design approach. The inverting configuration was used by Otala in this paper:
IEEE Transactions, Lohstroh, J. & Otala, M.: "An audio power amplifier for ultimate quality requirements"

Also:
http://home.online.n.../OtalaStory.htm

Pete B.

[Pete B]

>Hi Pete,
>
>I also find this very interesting and informative.
>
>Thanks,
>Doug

Thanks for reading, good to know there's interest!

Pete B.

[Doug G.]

Hi Pete,

I also find this very interesting and informative.

Thanks,
Doug

[Pete B]

One thing that I've noticed for years when listening to LP was the woofers flapping around due to, even mild, record warp and footsteps where the floor shakes a bit. I saw this recently at a show in NY in a room that has won best sound at the show. The woofers looked as if they were close to Xmax with the system playing at moderate levels. Interesting that the sound was very good, although it probably would have been better with a filter.

The second order high pass filter (Q108, Q109) in the Advent phono section completely solves this problem, there is no hint of subsonic garbage. I simulated this high pass filter in SPICE and found that the design includes a small amount of peaking and that the filter cutoff is set a bit high for an ultimate system. Obviously, this was designed for use in moderate cost systems.
The unity gain Salen and Key filter built around Q108 and Q109 has a peak of .9 dB at 28 Hz, is -3 dB at 15 Hz, -.1 at 19.7, -.5 at 18.8, -1 at 17.7, -10.3 dB at 10 Hz, -51dB at 1 Hz, and -90 dB at .1 Hz.

Increasing R122 and R123 to 68K, reduces the peaking to .09 dB, and the -1 dB is then about 18 Hz, and - 3 dB at 14 Hz.

Increasing C111, C113 from .1 to .22 uF divides these frequencies approximately in half, with -.3 dB at 10 Hz, -1 dB at 8.4 Hz, -3 dB at 6.4 Hz, -33dB at 1 Hz and -77 dB at .1 Hz. This offers a tight +/- .05 dB from 12 Hz to 20 kHz, yet offers very high attenuation at sub-sonic frequencies. I've not tried these revised values in the circuit.

I've not found most sub-sonic filters to be useful as they're often first order and set too high or too low.

Pete B.

[Pete B]

Thanks again Vern!

Listened to a few LPs the other night and the temperature has dropped again; the phono section was back to normal. It was an unusually warm night when it failed, I'll have to get out the soldering iron for heat and cold can for cold to test out the temperature theory on that transistor.

Pete B.

[dynaco_dan]

Hi Pete;

Interesting following your troubleshooting.

[Pete B]

After playing about half a side of an LP, one channel of the phono preamp began making a loud crackling sound followed by noise 20 to 40 dB higher than it should be. It's possible for the input diff pair transistors to become parially damaged over the years as a result of static discharge. This has been reported in other amplifiers and precautions can be taken in the circuit design to provide more protection. An anti static strap should be used when working inside electronic equipment.
Looked at the phono pre amp on a scope and it looks like very low frequency flicker noise and I was worried that it was the diff pair. It was not the line section since AUX in was fine. I shorted the node between C110 and C112 to ground to isolate the phono preamp from the emitter follower buffer expecting this to stop the noise. It did not, and therefore the problem is most likely the emitter follower transistor. This sounds like a transistor problem and it seems to be temperature related.

Pete B.

[Pete B]

The mute circuit can be tested by shorting, with a small value resistor to limit the discharge current, C222. The audio should mute while the cap is shorted and come back after one or two seconds after removing the short. This provides a test of the FETs.

Pete B.

[Pete B]

It is interesting that the diff amp transistor pairs are hand selected and matched. I've often wondered why more companies don't use dual transitors on a single die such as the LM394 which is actually an integrated circuit array of two groups of transistors in parallel. It is nearly an ideal transistor with low noise and nearly ideal matching being made from common silicon and operating at a common temperature.

Also interesting that the Advent uses NPN diff amp devices in the noise critical phono section, PNPs usually have lower noise.
It turns out that if transistors are bought in large lots, many will come from the same wafer and will therefore be near clones of one another. Matching is then easy and works well.

Pete B.

[Pete B]

Hi Vern,

Yes, relays on the output are common in high powered amps, but the more I think about it, the power amp topology here is a very common one, and it usually has better behavior. I'll have to think about what might be causing it. The relay is a bit complex for this simple design but I might use it if I had plans to use the power amp. Thanks for the suggestion.

Pete B.

[dynaco_dan]

>The unit works with the new caps, no surprises.
>Noticed printing SPI JUL 15, 1977 on the cover.
>
>Measured with a line voltage of 123.3 AC Vrms:
>Main raw +/- 22V supply at PSU caps: +25.6, -25.8V
>+/- 15 V reg: +15.2, -14.9V
>+/- 12V reg: +11.0, -10.7V
>
>These are fine the main supply is high because the line is
>also high.
>Looks like they might have lowered the +/- 12V to the phono
>section.
>
>Output stage bias reads .006V on .15 ohms or .006/.15 = 40 mA
>which
>is twice what it should be. Drove the amp with music to warm
>it up,
>measured again at the correct .003V on .15 ohms. The thermal
>feedback overcompensates, which is fine this is good for
>thermal stability.
>The unit warms up slowly with no signal, read the correct
>value when I measured it after a 2 hour warm up. Bias is not
>critical, and it starts out high so there's no risk of
>excessive crossover distortion.
>
>The serious power up thump is gone, however the power amp puts
>out a very low freq bump sound on power down, doesn't seem to
>be a serious problem. There's no way the mute circuit can fix
>this as it does it with the power amp disconnected from the
>preamp.
>
>Ended up recapping the entire amp including the power amp.
>Using it as a pre amp with a 100W/ch power amp, and it sounds
>fine. Even got out the old turntable.
>
>Pete B.


Hi Pete;

I am not a technical person, but I do have an idea that may be of use.

Add a line voltage or lower voltage 24V, 16V or 12V control relay DPDT, when the control AC is turned off, the relay drops out opening the speaker contacts.

Omron makes a very quiet relay model for example.

To add more work, Audio magazine had a delay or soft start circuit by, Walt Jung, which could pull in the speakers contacts after an adjustable slight delay and drop out instantly upon AC removal.

This soft start was for the surge bypass of the power supply of an amplifier through power resistors, but a speaker contact relay would be just fine as well.

This isn't what you wanted, more work, but what the heck, Heh, Pete. lol

You are doing a good job of resurrecting it, thank you.

By the way I don't own one of these units, I've missed a lot.

[Pete B]

The unit works with the new caps, no surprises.
Noticed printing SPI JUL 15, 1977 on the cover.

Measured with a line voltage of 123.3 AC Vrms:
Main raw +/- 22V supply at PSU caps: +25.6, -25.8V
+/- 15 V reg: +15.2, -14.9V
+/- 12V reg: +11.0, -10.7V

These are fine the main supply is high because the line is also high.
Looks like they might have lowered the +/- 12V to the phono section.

Output stage bias reads .006V on .15 ohms or .006/.15 = 40 mA which
is twice what it should be. Drove the amp with music to warm it up,
measured again at the correct .003V on .15 ohms. The thermal feedback overcompensates, which is fine this is good for thermal stability.
The unit warms up slowly with no signal, read the correct value when I measured it after a 2 hour warm up. Bias is not critical, and it starts out high so there's no risk of excessive crossover distortion.

The serious power up thump is gone, however the power amp puts out a very low freq bump sound on power down, doesn't seem to be a serious problem. There's no way the mute circuit can fix this as it does it with the power amp disconnected from the preamp.

Ended up recapping the entire amp including the power amp. Using it as a pre amp with a 100W/ch power amp, and it sounds fine. Even got out the old turntable.

Pete B.

[Pete B]

Your welcome Vern, nice to know your reading the post, thanks!

I noticed that the actual power amp in the receiver has resistors shunted by low value inductors for HF stability in series with the main output. This is not documented in any of the schematics, the old and new at that site, or even the printed schematic on the bottom of the unit.

Pete B.

[dynaco_dan]

Hi Pete;

Thank you for this very interesting and informative topic.

[Pete B]

The tuner on this unit has never worked right, it is the later design with ceramic IF filters. When it was given to me it sounded as if only the difference channel from the stereo decoder was coming out of L and R, and the signal was very weak. This was just a guess at the time.

I gave the pots a shot of DeOxit, repaired one of them that was broken, and recapped the tuner section. At this point the tuner had very weak output with lots of distortion, as if the signal was very weak. I became suspicious (because the output level control did nothing) of the amp (Q602 and Q603) between the FM demod and the Stereo Decoder, wondering if it had failed or might be oscillating. I accidentally found that if I touched the base of the first transistor (Q602) the signal became clear and strong. A cap to ground here worked also, this stage was probably oscillating. Ended up with 82 pF there and now the tuner seems to work fine. I would not normally just throw a cap at the problem, but I have a feeling that the design is flawed and don't want to spend anymore time on it. The tuner is working fine at this point so I'll leave the cap in for now.

Has anyone else seen similar tuner issues on the later ceramic filter version of this receiver?

[JKent]

I was given an Advent 300 Receiver many years ago by a coworker who said it was broken and not worth fixing. I offered to fix it but he wasn't interested. He said it thumped loudly on power up/down and something was strange with the tuner. I did nothing to it but used it on occasion as a spare amp.
I was never much interested in this amp as it does not have enough power for a main system, and the build quality is rather low.

I'm rebuilding it now just to have a working sample of the phono section, and will fix the line section so that it can be used as a full preamp.
No plans to work on the tuner or power amp at this time.

Verified operation, amp output offset low.
Power supply voltage good.
Unit pops badly on power up/down.
Some controls are noisy.

1. Sprayed all contacts with Deoxit
2. Replaced pair of 3000 uF (date code 22 week of 1977) power supply caps with a pair of 10,000 uF.
3. Replaced all line section PSU electrolytic caps with tantalums, I do not normally suggest these but I had some good quality hermetically sealed units.
4. Replaced all Line signal path caps with GE poly types.
5. There is no tone control defeat, hard wired tone controls out of circuit.
6. There is a FET based mute circuit on the output of the line preamp to avoid power up thumps and pops. This was not working because the cheap electrolytic timing capacitor was far out of spec, if not completely shot. Replaced with a tantalum and at least the most offending pops are gone.

Some comments based on the schematics:
1. Interesting that the line stage is inverting.
2. Volume pot is 25K ohms, which is low and approximately defines the line level input impedance.
3. Volume pot drives the tone control section. The line level input impedance is flat with tone controls set flat, but becomes frequency dependent with non-flat tone settings. I verified this through SPICE simulations.
4. Pre amp power supplies employ active regulation, line +/-15V, phono +/-12V.
5. Line stage is a primitive transistor based OP amp with a simple diff amp and VAS employing 3 transistors.
6. There is no RF filtering, line stage is -3dB at 1Hz and 490 kHz as predicted through simulation.
7. Phono section utilized Holman's high pass filter at output of preamp rather than the input as described in the AES article.
8. Split supplies are used with OP amp type amplifiers which could offer direct coupled signal paths if attention was paid to DC offsets, however capacitor coupling was used instead.
9. PC board layout is haphazard making it difficult to locate parts.

Pete B.

Hi Pete
I just bought a 300 to play with (like I don't have enough unfinished projects here )
So you recommend replacing the 3000uF 30vdc PS caps with 10,000uF? What voltage? The original Spragues are about 1" x 2" axials. Hard to find something that will fit physically in that high value. On this page is an 8,000uF 35v axial, 1" x 2 5/8"
http://www.surplussa...uF-99999uF.html

assuming the length will make it, does that look appropriate? Fortunately, the Advent seems to be working OK on 1st listen, but I thought replacing the caps might be appropriate.
Thanks for any input
Kent

[Pete B]

Hi Pete
I just bought a 300 to play with (like I don't have enough unfinished projects here )
So you recommend replacing the 3000uF 30vdc PS caps with 10,000uF? What voltage? The original Spragues are about 1" x 2" axials. Hard to find something that will fit physically in that high value. On this page is an 8,000uF 35v axial, 1" x 2 5/8"
http://www.surplussa...uF-99999uF.html

assuming the length will make it, does that look appropriate? Fortunately, the Advent seems to be working OK on 1st listen, but I thought replacing the caps might be appropriate.
Thanks for any input
Kent

Hi Kent,

10,000 uF was just what I had on hand, 3000 or up will do as long as they fit and
meet the voltage spec.
The rest of the electrolytics are very cheap units and were going bad. The polys
I used are large, and they have to be elevated to fit - doesn't look very neat.

[JKent]

1. Sprayed all contacts with Deoxit
2. Replaced pair of 3000 uF (date code 22 week of 1977) power supply caps with a pair of 10,000 uF.
3. Replaced all line section PSU electrolytic caps with tantalums, I do not normally suggest these but I had some good quality hermetically sealed units.
4. Replaced all Line signal path caps with GE poly types.
5. There is no tone control defeat, hard wired tone controls out of circuit.
6. There is a FET based mute circuit on the output of the line preamp to avoid power up thumps and pops. This was not working because the cheap electrolytic timing capacitor was far out of spec, if not completely shot. Replaced with a tantalum and at least the most offending pops are gone.

Pete B.

Hi Pete
Flush with enthusiasm after recapping my MAC4100, I'm turning to the Advent 300. Two questions right off the bat, as I put my Mouser order together:

• That's quite a jump from 3,000uF to 10,000. I understand that's what you had on hand. Is there any benefit to going much higher than 3,000? And what voltage? Originals are 3000/30. Mouser has several that will fit, physically: 10,000/50, 6800/63, 5600/100 3300/35 etc etc etc. As long as I'm ordering, what would you consider "ideal?"
  
• Will a big change in those caps require other changes?

Generally I plan to use all audio-grade caps of the same capacitance but one step up in voltage (in most cases).
Oh--one other question. C104 and C105 are listed on the schematic as 4.7/10 tantalum. Should they be replaced with new tantalums, electrolytics, NPE, or film?
Thanks for any tips
Kent