Grundig Amp for Spring Reverb

[ERRP55]

So if I am going with the 4:1 stepdown transformer (CMLI-10-600) is it safe to not worry about the 100mV max input source warning on the schematic? Does the voltage divider that was shown in @rock soderstrom comment not need to be present to have this spring reverb working properly with the signal I will be feeding the triode? Also, because I will be removing r24 (1M pot) from the output does the CMLI-10-600 transformer REVERSED work for the output? I am removing the 1M pot because the reverb will show up on a channel return on the mixer or the DAW. I can use a fader adjust level. @Brian Roth @metalockpick THANKS!

 

[metalockpick]

Removing 1 M pot- Sure. Putting a "Tone" (high cut) pot before the last gain stage would help to insure you don't clip the last stage, and also to change the reverb voicing. You could use a dual pot and make the other section a series resistor that would be minimum resistance @ full tone cut and maximum resistance @ full brightness to compensate for the level loss @ low "tone" settings.
I've included the fender (reissue) schematic of their first stand alone reverb; everything up to the "mix" control is relevant. The "Drive" control there affects the character of the sound by setting how hard you hit the springs...

Yes, the 4:1 transformer would be used as a stepdown on the output as well (10k side to the tube, 600 Ohm side would be the output). That gives you an isolated and balanced output.

Reverb is typically around 300 Hz to a few kHz. You don't want any output in the bass range, as phase cancelation will play havoc with your low end.

Hitting the springs with as low as say, 80 Hz will give you a crashing/boinging effect, which may be desirable; the output from the springs is likely going to be 3rd harmonic and up, so figure 240 Hz+ (80x3) will be your output frequency range.
A rotary switch to set the HPF by selecting different capacitor values to roll off bass before the spring drive amplifier will make it more fun to use by emphasizing or de-emphasizing certain frequencies. Cap values? As the chef's say, to taste...

 

[ERRP55]

Removing 1 M pot- Sure. Putting a "Tone" (high cut) pot before the last gain stage would help to insure you don't clip the last stage, and also to change the reverb voicing. You could use a dual pot and make the other section a series resistor that would be minimum resistance @ full tone cut and maximum resistance @ full brightness to compensate for the level loss @ low "tone" settings.
I've included the fender (reissue) schematic of their first stand alone reverb; everything up to the "mix" control is relevant. The "Drive" control there affects the character of the sound by setting how hard you hit the springs...

Yes, the 4:1 transformer would be used as a stepdown on the output as well (10k side to the tube, 600 Ohm side would be the output). That gives you an isolated and balanced output.

Reverb is typically around 300 Hz to a few kHz. You don't want any output in the bass range, as phase cancelation will play havoc with your low end.

Hitting the springs with as low as say, 80 Hz will give you a crashing/boinging effect, which may be desirable; the output from the springs is likely going to be 3rd harmonic and up, so figure 240 Hz+ (80x3) will be your output frequency range.
A rotary switch to set the HPF by selecting different capacitor values to roll off bass before the spring drive amplifier will make it more fun to use by emphasizing or de-emphasizing certain frequencies. Cap values? As the chef's say, to taste...

My thought was to use Channel EQs to shape the tone once I return the signal back to the board. But i do like the idea of using a dual pot for the tone control between the last gain stage. I was looking at using a Greyhill/Elma style rotary switch for selectable hi pass filters at the junction of c12 and r12. I was thinking frequencies around 90hz, 150hz, and 300hz. I'm going to redraft my schematic and post it tonight. Thanks letting me pick your brain!

 

[metalockpick]

Channel EQ- Of course you will.
The TONE control is going to be a different curve than channel EQ...It's a bonus feature that may come in handy (because you can really sink the HF with those things). It won't otherwise detract from the performance.
A Lorlin switch can be easily set for the number of intended positions, and a single pole/deck is all you need to switch between capacitors. Grayhill is a nice switch, but a ham-handed operator (me) can destroy the switch stops on those without much effort.
Large resistors (1-5 meg, pick one) between the caps can eliminate pops between settings (not really an issue, as you wouldn't likely be switching with the effect up).

 

[nsalis]

Let's clarify a few premises:

You want to provide the Grundig HS-1/HV-1 combination with transformer-balanced inputs and outputs, both with XLR connectors. This suggests that you want to use the device in a studio environment. In such an environment, XLR means the following:
1. Input: Impedance 10k Ohm or higher, nominal level (0 VU) = 4 dBu (1.23 V), maximum level without distortion at least = 18 dBu (6.15 V).
2. Output: Impedance 1k Ohm or lower, nominal level (0 VU) = 4 dBu (1.23 V), maximum level without distortion at least = 18 dBu (6.15V).

So we assume a so-called headroom of (18 - 4 =) 14 dB.

The Grundig HV-1/ HS-1 is a valuable technical witness of its time. We should leave it intact, i.e. not alter it and make additions outside of it.

You obviously own a Transformer Cinemag CM-13102 that you want to put to good use. This transformer has a primary winding of 10k Ohm and two secondary windings of 10k Ohm each. (That is, it is NOT a "high impedance to line transformer" as you put it. It is a transformer whose primary winding and secondary windings are of the same type. For this reason, the level and impedance ratios are not changed during the transformation.)

The Grundig device can accept a maximum of 100 mV (-18 dBu) at the input before it starts to distort. If you subtract the headroom of 14 dB from these -18 dBu, you end up with (-18 minus 14 =) - 32 dBu (10 mV). This corresponds to a level that microphones emit.

If you want to use the C13102 to convert a balanced 4 dBu level to an unbalanced -32 dBu level, you need a voltage divider between the secondary winding of the input transformer and the input of the Grundig amplifier. The easiest way is to use a potentiometer. This is a variable voltage divider that allows you to drive the input of the amplifier just far enough that it doesn't start distorting. I enclose a sketch of the circuit.

It is somewhat more difficult to adapt the output of the Grundig amplifier to studio conditions. With the volume control R24 turned all the way up, we have an unbalanced signal that can be as high as 28V and expects an input impedance of the subsequent device of at least 100k ohms. For use in the studio we are now looking for a converter that has a high-impedance input (greater than or equal to 100k ohms), a low-impedance output (less than or equal to 1000 ohms).

A passive transformer would meet the requirement if it could step down the impedance from 100k ohms to 1k ohms. This transformer would have a turns ratio of 10:1 and step-down the voltage from 28 Volt to 2.8 Volt . Quality line level transformers like this are hard to find. That's why I prefer an active device à la Sonifex RB-Li2.

 

[metalockpick]

...also, never draw outside the lines, don't question your elders, and turn down that awful music

 

[ERRP55]

Let's clarify a few premises:

You want to provide the Grundig HS-1/HV-1 combination with transformer-balanced inputs and outputs, both with XLR connectors. This suggests that you want to use the device in a studio environment. In such an environment, XLR means the following:
1. Input: Impedance 10k Ohm or higher, nominal level (0 VU) = 4 dBu (1.23 V), maximum level without distortion at least = 18 dBu (6.15 V).
2. Output: Impedance 1k Ohm or lower, nominal level (0 VU) = 4 dBu (1.23 V), maximum level without distortion at least = 18 dBu (6.15V).

So we assume a so-called headroom of (18 - 4 =) 14 dB.

The Grundig HV-1/ HS-1 is a valuable technical witness of its time. We should leave it intact, i.e. not alter it and make additions outside of it.

You obviously own a Transformer Cinemag CM-13102 that you want to put to good use. This transformer has a primary winding of 10k Ohm and two secondary windings of 10k Ohm each. (That is, it is NOT a "high impedance to line transformer" as you put it. It is a transformer whose primary winding and secondary windings are of the same type. For this reason, the level and impedance ratios are not changed during the transformation.)

The Grundig device can accept a maximum of 100 mV (-18 dBu) at the input before it starts to distort. If you subtract the headroom of 14 dB from these -18 dBu, you end up with (-18 minus 14 =) - 32 dBu (10 mV). This corresponds to a level that microphones emit.

If you want to use the C13102 to convert a balanced 4 dBu level to an unbalanced -32 dBu level, you need a voltage divider between the secondary winding of the input transformer and the input of the Grundig amplifier. The easiest way is to use a potentiometer. This is a variable voltage divider that allows you to drive the input of the amplifier just far enough that it doesn't start distorting. I enclose a sketch of the circuit.

It is somewhat more difficult to adapt the output of the Grundig amplifier to studio conditions. With the volume control R24 turned all the way up, we have an unbalanced signal that can be as high as 28V and expects an input impedance of the subsequent device of at least 100k ohms. For use in the studio we are now looking for a converter that has a high-impedance input (greater than or equal to 100k ohms), a low-impedance output (less than or equal to 1000 ohms).

A passive transformer would meet the requirement if it could step down the impedance from 100k ohms to 1k ohms. This transformer would have a turns ratio of 10:1 and step-down the voltage from 28 Volt to 2.8 Volt . Quality line level transformers like this are hard to find. That's why I prefer an active device à la Sonifex RB-Li2.

I actually do not own a CM-13102. I mentioned this transformer because it is a balanced to unbalanced transformer, I confused it with their 4:1 step down transformer (CMLI-10/600).

However thank you for bringing up the issue of the output stage. To combat the possibility of a seeing a 28v output and the need for the HV-1 to see a 100k input source. Can a calculated voltage divider go in place of the 1M pot to attenuate the signal going into the 4:1(CMLI-10/600) step down TX or we into a JD-DB-E transformer? The JD-BE-E is a 12:1 step down however it is labeled as unbalanced to mic level. But considering the output that the amp will be driving, once the signal has passed thru the 12:1 transformer wouldn’t it still be line level?

No trying to shoot down the idea of an active device “Sonifex”, apologies if my questions and comments are repetitive and/or foolish. I’m just trying to make sure I understand everything as some concepts are still foreign to me. Doing my best to learn from you veterans. Thanks!

 

[ERRP55]

...also, never draw outside the lines, don't question your elders, and turn down that awful music

You have my word for the first 2 but my music taste is not awful!

 

[metalockpick]

1) I've used both 4:1 and 12 to 1 cinemag (I like them because they're high quality, ethical and local LA) transformers for this kind of circuit. 12:1 get used when you're deliberately overdriving the tube ("28V"). 4:1 is what you use when you're running it clean.
2) a quality transformer sounds better than a resistive voltage divider, every time.
3) All of the numbers and word salad don't articulate what it's going to sound/perform like and why, and take no account of your abilities as an engineer.
4)I should have added "YOU KIDS GET Off my LAWN" to the last post.

 

[nsalis]

I actually do not own a CM-13102. I mentioned this transformer because it is a balanced to unbalanced transformer, I confused it with their 4:1 step down transformer (CMLI-10/600).

However thank you for bringing up the issue of the output stage. To combat the possibility of a seeing a 28v output and the need for the HV-1 to see a 100k input source. Can a calculated voltage divider go in place of the 1M pot to attenuate the signal going into the 4:1(CMLI-10/600) step down TX or we into a JD-DB-E transformer? The JD-BE-E is a 12:1 step down however it is labeled as unbalanced to mic level. But considering the output that the amp will be driving, once the signal has passed thru the 12:1 transformer wouldn’t it still be line level?

No trying to shoot down the idea of an active device “Sonifex”, apologies if my questions and comments are repetitive and/or foolish. I’m just trying to make sure I understand everything as some concepts are still foreign to me. Doing my best to learn from you veterans. Thanks!

The JT-DB-E (not JD-...) is a very fine product. (I've specified a very large amount of Jensen transformers over the last 30 years and never had the slightest defect.)

The JT-DB-E can accept a maximum of 21.5 dBu (9.2 volts) without distortion. Your Grundig can emit a maximum of 28V (31.2 dBu). If you want to use the Jensen, you would have to use the Grundig's output level control to reduce the level so that - taking into account the necessary headroom of 14 dB - a maximum of (21.5 - 14 =) 7.5 dBu nominal level (0 VU) is sent. The Jensen transformer lowers the level by 22 dB from input to output which results in an output level of (7.5 - 22 =) -14.5 dBu (145 mV). That is about 20 dB lower than a normal studio level (1.23 V = 4 dBu) or about 6.5 dB lower than consumer gear level (0.316 V = -10 dBV = -8 dBu). If you can live with that, i.e. if you can feed the signal into the microphone input of a mixer to use it further, then it's ok. However, if you expect a standard studio level of 0VU = 4 dBu (1.23 V) at the converted output of the hall spiral amplifier, the solution with the Jensen stepdown transformer JT-DB-E will not work and you need the active device I mentioned.

Nick Salis

 

[metalockpick]

The RE12-DBE / JT-DBE / whatever Cinemag calls it now is a 12:1 transformer that is an Ed Reichenbach design. Ed Reichenbach designed and manufactured all of the original Jensen transformers. Reichenbach Engineering, all of its designs and all of its equipment became Cinemag. Reichenbach designed transformers for Altec in the '40s and '50s, then went solo. his audio transformers were the choice for high end West Coast studio and sound reinforcement for decades; Jensen was his marketing arm. The split was acrimonious, and now there are two companies.

The DBE does some very interesting sonic things when it is not properly terminated(loaded) that people do seem to like.
That original design transformer is also available from Cinemag.
It is inappropriate here because you are not going to be slamming the front end of that reverb unit. A 4:1 is sufficient. Also, there is no reason to have the transformer input or output balanced with reference to ground, isolated from ground is preferable, and the transformers will give you that.
Preserving that unit for future generations is stupid, because adding reverb and noise to mixed and mastered sound with a consumer grade product was a terrible idea to begin with. Repurposing it for a unique studio effect is a great idea.

 

[rock soderstrom]

The Grundig HV-1/ HS-1 is a valuable technical witness of its time. We should leave it intact, i.e. not alter it and make additions outside of it.

I don't see it that way, as the unit will spend its future existence in a studio environment, it's absolutely fine to modify it for that.

My two €-cents about the transformer in this application. I don't see the need for high $$$ quality iron here, the signal is so degraded by the spring, there is no advantage in high end transformers. Listen to the wet signal solo and you will realise that the spring and the transducers make the sound.

Galvanic isolation at the input with a suitable transformer makes sense. As already mentioned, I would adapt the circuit a little. At the output I would establish a CF.

For me, the most important things are an adjustment option for the level with which the spring is fed and how the whole thing sounds in practical use. I would start with this and make any modifications dependent on it.

cheers

 

[nsalis]

By the way - we haven't even talked about what happens if you leave the reverberation device in the Grundig console and use it that way. As far as I know, it's a Grundig SO390. This is an exclusive product and was very expensive at the time. Provided that at least in the audio part still works and you have space for it in your studio, you could connect the unit as follows:

The adapter cable from the loudspeaker output of the Grundig console (a DIN loudspeaker socket, which, depending on how the plug is inserted, lets the built-in loudspeakers continue to run or switches them off) is to be wired as follows with the XLR 3m plug:

Round contact = XLR pin 2, flat contact = XLR pin 3. XLR pin 1 remains free. Any 2-wire cable (unshielded) can be used.

If you insert the plug in such a way that the built-in loudspeakers do not run, you should connect a 15 ohm/ 5 watt resistor between XLR pin 2 and XLR pin 3 in order to protect the valuable output transformers. This resistor is physically too large to accommodate it inside the plug. The loudspeaker output is a very good driver for the 4 dBu input that your studio gear has.

In order to route a balanced 4 dBu studio signal to the console's reverberation unit, you would have to implement the circuit I drew with the 1:1 transformer and potentiometer. The wiper of the potentiometer would go to pin 3 and pin 5, the shield would to pin 2 of the TA socket (pick-up = phono input). Before doing this, unplug the turntable plug that is plugged into this socket.

If you are not familiar with the pinout of these DIN speaker and phono plugs or you do not have access to such connectors, I can help.

If you now switch the Grundig console to TA, the incoming signal will be audible and can be provided with reverberation.

The volume control know on the Grundig Console must be about 1/2 open. Don't turn it up too high or the power amplifier will start clipping. You have to play with the sound controls on the console. In general, such consoles have too much bass. To start with, I would turn off all the bass and turn the treble all the way to the right.

 

[nsalis]

I don't see it that way, as the unit will spend its future existence in a studio environment, it's absolutely fine to modify it for that.

My two €-cents about the transformer in this application. I don't see the need for high $$$ quality iron here, the signal is so degraded by the spring, there is no advantage in high end transformers. Listen to the wet signal solo and you will realise that the spring and the transducers make the sound.

Galvanic isolation at the input with a suitable transformer makes sense. As already mentioned, I would adapt the circuit a little. At the output I would establish a CF.

For me, the most important things are an adjustment option for the level with which the spring is fed and how the whole thing sounds in practical use. I would start with this and make any modifications dependent on it.

cheers

I understand your point of view, but please understand me: I am a collector and restorer of such devices, and it hurts my heart when such treasures are destroyed and cannibalized.

 

[Tubetec]

Variable resistive attenuation is what you want beween the drive output and the spring , you want to be able to introduce some distortion in the drive electronics without totally oveloading the spring drive transducer .
I have a Peavey Nano Valve 5w amp , I rigged an extra attenuated output from the speaker with a 200 ohm W/W pot , that allows me all the scope I need to trim the drive , reverb recovery is done seperately and the reverb itself is reproduced over monitors , not fed back through the amp .
Keeping the reverb recovery stage as low noise/low distortion as possible generally improves matters where your looking for a less coloured sound .

 

[rock soderstrom]

I understand your point of view, but please understand me: I am a collector and restorer of such devices, and it hurts my heart when such treasures are destroyed and cannibalized.

I know exactly what you mean, I have already restored a Grundig "Klangtruhe" myself and still have two Siemens ones in my cellar, along with a couple of tube R2R recorders and other great stuff.

Unfortunately, these devices also have a very low resale value and take up an extremely large amount of space.

Which sadly leads to the current reality, at least here in Germany. These devices are simply thrown away every day.

That's why I don't see it as cannibalisation, but rather as the preservation and use of valuable resources. Everything is much better than throwing them away or incinerating them.

 

[ERRP55]

@metalockpick @nsalis @rock soderstrom Thanks for all of your input and suggestions. I restore grundigs, telefunkens, and Philip stereo consoles as well. I always do my best to keep them all original (electronics, turntables, speakers etc.) the only reason why I took the spring reverb out of the SO 390 was because it’s not going to be used. It is definitely a cool feature but it will sit in the cabinet collecting dust. I’ve sold and fixed a few units so far that have built in spring reverbs and most if not all folks don’t understand it and never use it.

I’ve drawn up a schematic with all of the modifications that you all suggested. It’s split into to photos (pre and post spring reverb tank) Please let me know if I missed anything or if I should make changes. Apologies for the sloppy hand writing.

 

[rock soderstrom]

I have only glanced at it briefly. The 250k dwell potentiometer suppresses the gridleak biasing through the 10M resistor, so that won't work.

 

[ERRP55]

I have only glanced at it briefly. The 250k dwell potentiometer suppresses the gridleak biasing through the 10M resistor, so that won't work.

I’ll eliminate that from the drawing. the only reason I included it was to combat a potential rumble noise that may occur from the dwell pot to the grid. Is the grid leak resistor even needed there? Would it better to have it placed after the Hi-Pass Filter before pin 8 (grid of pentode)?

 

[rock soderstrom]

I’ll eliminate that from the drawing. the only reason I included it was to combat a potential rumble rf(noise) that may occur from the pot to the grid. Would it better to have it placed from the junction of r12 to pin 8 (grid of the ECL86 triode)?

I think you mean right before the pentode?

I don't understand the rest of your statement either. Sorry.

I think you need to clarify for yourself if you want to keep the grid leak biasing or not. I have written my opinion about it.

Just try what you want to do and you will see what works and what doesn't...