BAJA CV30M1960A Speaker Simulator pedal

bajaman · Post by **bajaman** » 07 Apr 2018, 08:14

Hey folks - here is something different for you to consider NJACS (not just another cab sim!)

A few years back I was playing around with cabsims for direct recording to a mixer or soundcard with no guitar amplifier. Most all the designs I tried including the Marshall SE100 and the infamous LXH2 designs left a lot to be desired - they all use lots of op amp filters that ended up making the guitar sound flat and lifeless in my opinion. These days cheap dsp has meant impulse response based designs work a lot better, but you should know me by now - I like to do things a little differently. Sometimes they work - sometimes they end up stashed away in a dark corner of the workshop

Anyway being a diy kinda guy and for something to do, I revisited an older design of mine in the hopes of improving it and creating a working pedal - here is the result of my efforts, a speaker cabinet simulator based on the response characteristics of the Marshall 1960A quad box loaded with Celestion vintage 30 loudspeakers. In addition I have included the response characteristic of a typical tube output stage, output transformer reactive loading stage with the cabinet response - this can be switched in or out for comparison

Please note: THIS IS NOT A LOAD BOX FOR USE WITH THE OUTPUT OF A VACUUM TUBE OR HIGH POWER SOLID STATE AMPLIFIER
This pedal is designed to be used as the final link before a mixer or soundcard with NO POWER AMPLIFICATION required

I have included a balanced XLR DI output and a switchable true bypass unbalanced jack output. The jack output has a level control.
Anyway here it is:
enjoy

bajaman

temol · Post by **temol** » 07 Apr 2018, 08:25

Any chance for switchable mic positions? Something like center, cone, edge.

T.

graeme · Post by **graeme** » 07 Apr 2018, 09:00

You're my hero, Bajaman!!!!!

287m · Post by **287m** » 07 Apr 2018, 09:50

look at schemo. where the dpdt?

then look at guts, oh man, just spdt.

hahaha

btw, how to change to G12M greenback?

====
maybe this is silly Baja

just for portability, ever try design like this
amp in a box/preamp -> cabsim -> (out1) mixer, (out2) headphone with aux in?
the request came from friend who travel a lot to remote area, where he can do the job very fast and get bored play acoustic, when want play electric, the electricity not always 24 hours on!
and yes, as a LP guy, he cant afford like this

bajaman · Post by **bajaman** » 07 Apr 2018, 11:23

typo on schematic corrected

The gutshot shows a slightly different layout to the rev 1 vero, but is the same electronically

Also I used OPA2134 because I did not have any of the far cheaper TL072 op amps in stock - either of these or NE5532, LM833 etc will work just fine - hell, you could even use low current TL062 devices if you plan on using battery power

And yes, I used a blue led not an orange one for the status light with a 33k current limiting resistor (not 1k5 as shown on schematic)

cheers
bajaman

temol · Post by **temol** » 07 Apr 2018, 13:10

I have it on a breadboard now, working. I'll try to capture some riffs later.

T.

temol · Post by **temol** » 07 Apr 2018, 20:30

No sound samples yet. As for now I can post frequency response graph. There are four plots. With number 1 - 1uF multilayer ceramic capacitors, with number 2 - 1uF multilayer polyester capacitors (dipped type, MKT).

T.

bajaman · Post by **bajaman** » 07 Apr 2018, 21:27

temol wrote:No sound samples yet. As for now I can post frequency response graph. There are four plots. With number 1 - 1uF multilayer ceramic capacitors, with number 2 - 1uF multilayer polyester capacitors (dipped type, MKT).

[ Image ]

T.

That preamp out plot with the film capacitors looks wrong in the low end - are you sure you are not using the wrong value parts in the opamp #4a section ?

Thanks for the plots though they do show the response I was trying to achieve - I had to lower the tube driven output section level (potential divider with two 4k7) to match the output level of the preamp out and to prevent clipping the filter stages.
By the way - higher voltage supply rail is always beneficial - 18v up to 30v

Looking forward to your sound clips temol

cheers
bajaman

bajaman · Post by **bajaman** » 07 Apr 2018, 21:52

Here are some Multisim plots for the pre out frequency and phase shift and then the tube driven output frequency and phase shift. i cannot understand why temol's plots show such a high peak at 5kHz as these theoretical simulation plots certainly do not show it

temol · Post by **temol** » 07 Apr 2018, 22:11

I'll recheck values and redo the breadboard layout. Check the recording on souncloud. That's all I was able to record today, nothing fancy.

No cab, then four times BAJA CV30M1960A, then cab IR two times.

T.

sixthfloor · Post by **sixthfloor** » 08 Apr 2018, 10:02

To my ear there is definitely a high peak in the recordings of the Baja CV30M1960A compared to the IR. Thanks for sharing, Temol

bajaman · Post by **bajaman** » 08 Apr 2018, 10:57

Well all I know is that this design uses two sallen key slightly underdamped low pass filters ( the second of which is slightly modified to include a small mid peak response around 1kHz).
The filters each have the following characteristics:
Cut-off frequency
fc = 5616.651827034[Hz]
Quality factor
Q = 1.4398509407947
Damping ratio
ζ = 0.34725816807401

From this calculated data one would expect to find an approximately 3dB peak in the response around the cut off frequency - a maximum of +6dB with both filters in series

Certainly no more than 9dB, so i am at a complete loss to understand how temol gets +15dB peaking at just over 5kHz

temol · Post by **temol** » 08 Apr 2018, 11:28

bajaman - can you test your real circuit with rmaa?
I've just finished re-breadbaording and measuring the sim. Same story.. strong 5kHz peaks.
As for the plot difference between film and mlc caps. It may be caused by different capacitance values. Film caps - 1uF, mlc around 930nF (tested with cheap LCR meter).
Tube driven output is louder than pre - around 2-3 dB.

T.

sixthfloor · Post by **sixthfloor** » 08 Apr 2018, 16:59

287m wrote: btw, how to change to G12M greenback?

I think the G12M <-> V30 conversion was asked back when Bajaman published his initial circuits based on the celestion blue and the M1960A.
https://www.freestompboxes.org/viewtopic ... 13&t=22857
https://www.freestompboxes.org/viewtopic ... 3&p=226062

From what I gathered from these topics :
- change the 270n cap near IC2a to 220n;
- change the 27k resistor near IC4b to 18k.

Note this was said for the initial M1960A design. Comparing the schematics there are many similarities between them but the circuits are not identical, hence these changes might not give the expected result.

bajaman · Post by **bajaman** » 09 Apr 2018, 01:17

some quick and dirty plots from the actual finished pedal

theory and practice are in agreement here

black trace is preamp out, blue trace is power tube driven out

cheers
bajaman

: Image1.png (2.05 KiB) Viewed 6796 times

bajaman · Post by **bajaman** » 09 Apr 2018, 05:48

same graph but more frequency steps

: Image2.png (2.08 KiB) Viewed 6785 times

temol · Post by **temol** » 09 Apr 2018, 06:21

So... how about difference in opamps? I'm using TL072.

I attach a frequency response graph I've taken from another cabsim (in the past). Four different opamps.

T.

bajaman · Post by **bajaman** » 09 Apr 2018, 10:30

No - all the simulations were with TL072 and the pedal is using 4 x OPA2134 devices, but i very much doubt that would make any difference here. My guess is something wrong on your breadboard - build the vero layout and compare perhaps

By the way - I took the pedal around to my jam session with my friends earlier tonight - I gave it to one of the guitarists, who usually uses a string of Boss pedals straight into the mixer / soundcard input - i placed the pedal at the end of the chain and it stayed on that setting (tube out) all night - very fat bottom end (good on the bottom E string

) and silky top end - impressed so far - will see how it goes next week, but i definitely felt it improved the focus of the guitar sound - definitely has that tube amplifier bottom end and top end sparkle

cheers
bajaman

temol · Post by **temol** » 10 Apr 2018, 18:25

I cannot locate 1,5M input resistor on a vero and on the photo of the board. Is it ommited or just hidden somewhere?
There is also 6k8 instead of 5k6 on the vero layout.

T.

bajaman · Post by **bajaman** » 10 Apr 2018, 22:00

temol wrote:I cannot locate 1,5M input resistor on a vero and on the photo of the board. Is it ommited or just hidden somewhere?
There is also 6k8 instead of 5k6 on the vero layout.

T.

The 1M5 is not entirely necessary - so, yes it is omitted from the pedal and vero layout. It is only there to minimize dc switching pulses when engaging the bypass switch, but the input capacitor is grounded when in bypass anyway so i decided to leave it out. It makes absolutely no difference to the frequency or phase response of the pedal

Yes - good spotting

the 6k8 that connects to the two 33n capacitors in the mid notch filter section should indeed be a 5k6, but again this mistake on the vero and my finished pedal only makes a slight (+ .75dB ) boost in the mid response centred around 1kHz or thereabouts. Neither of these two differences explain the highly underdamped peak in response at 5kHz that your plots show - what are you using to plot these curves

I am using a Vellemann PCSUG250 (https://www.velleman.eu/products/view/?id=377622) running their PcLab2000LT software and it shows almost identical curves to my simulated design in National Instruments multisim.
Anyway here is the corrected vero layout for all - thanks temol

for noticing the mistakes.
cheers
bajaman