Building the boards, PT 1

Today I began the task of building the turret populated boards that will hold the amplifier circuitry. I have decided to omit the relay-switching of channels that I had planned and opt for the simpler "always-on" mixing circuit that the original Fenders had. Using this method I can still have reverb on the solid-state channel by jumping out of 1 of the unused inputs into the tube Fender channel, and be able to mix it in as necessary.

Turrets - what are they?

They are used as junction points for circuit connections. The device is installed onto a board - fiberglass, in this case, by a special turret staking tool. Once installed, components can be soldered directly to the turret, thus making an electrical connection. It can be considered one method of point-to-point wiring.

In the old Fender amps, they used solder "eyelets" which were essentially contact points like the turrets, but had a profile that was nearly even with the board level.

I chose turrets because I have some, and because a mechanical connection can also be achieved with them, strengthening the connection of the components.

I began the design of the turret board layout by using a CAD program called CADStd. The designer of the software offers a freeware version of the program than is very user friendly and has a quick learning curve.

Using CADStd, I drew up several of the components I had that would be going into the build of the tube amp - then after drawing them up, stopped. It became clear to me that things would go much quicker if I simply worked with the actual components, on the actual fiberglass board, and draw directly on the board.

This decision to head away from the computer was a good one. I took my schematics and layouts and laid everything I had out on the kitchen table and began setting things up. Within a few hours I had completed the layout for the turret board.

Now that I have this done, I need to drill the board (including holes for wires), install the turrets and begin populating the board.

I also have a self-etched PCB of the solid state preamp that needs preparation.

Regarding the turret staking tools, I may have to build the part that presses against the top of the turret, as I do not have this part, and none of the Keystone staking tools do not seem to match the size of the turrets that I have.

2:50PM-7PM; 11PM-12AM (5.17 HRS)

Back on the project. Parts ordering.

Probably the most boring part of the project, but someone's got to do it. I have a method of compromise between price, quality, and stocked items from a distributor. I've gotten pretty good at getting what I believe to be the lowest price for parts (at this quantity level).

The hard part is trying to remember if I need any parts for other projects. Now's the time to buy, so I can save on shipping.

Another somewhat time consuming part is finding parts that fit dimensionally within the scope of my project. If you search for 22uF on Mouser's site, you see that there are literally thousands of 22uF capacitor types that can be purchased. So it's a game of switch back and forth between schematic, website, price guide, dimensions, and repeat.

It's really worth it to me in the end since I want to spend the least amount of money on parts (while getting a quality product). Doesn't everyone want this?

10PM-12AM (2 HRS)

Began parts-hunting for the project. Despite the amount of parts and miscellaneous hardware I have in my garage, I am missing a fair share of components. I have the heavy stuff - that counts for quite a bit actually.

I created a spreadsheet in Open Office's "CALC" program - an open source version of a well-known corporate company's EXCELlent program. Then, while uploading the file to google, it dawned on me - why not just have the spreadsheet online? I remember seeing some time ago that google docs can do just that - I uploaded my Open Office file and google converted it to it's online version. Now I can make changes from any computer and not worry about which PC at home had the most recently modified spreadsheet. Pretty cool.

Here's the link:

https://docs.google.com/leaf?id=0AkSr64f9Rb3DdFpQMHFCa2JlVE1zTVhfclNiSkxCVHc&sort=name&layout=list&num=50 

And an embedded html view of the spreadsheet:

I am going to wait on some parts and hope I can make a trade-able donation to SCC's Electronic Student Association for some small passive components

10:00AM-2:00PM (4HRS)

Today I finished drawing up the circuit sections for the tube amplifier that I intend to build. From this, a part list shall be made, and parts acquired. I ran some quick simulations of the preamp...

I successfully modeled a simple center tapped output transformer as well as the reverb driver transformer; based on the LTSpice manual, and examples on the web. The modeling created represents a simple, ideal transformer model. There are many parameters (beyond the scope of this class, i.e. engineering degree work) that should be considered if a true representation of a transformer is desired/required.

I also reinforced some electronics theory while working today:

Impedance ratio vs. Voltage/Turns ratio

Ratio(Z) = [Ratio(turns)] * [Ratio(turns)]

The output transformer I intend to use is rated (not verified) for 7.5k:8 OHMS.

The voltage/turns ratio is calculated as such:

That is, the turns ratio is 30.619 : 1

I don't know if I will have time to spec out the output transformer, but it would be interesting to document the parameters of the transformer, and compare actual performance versus simulated performance. I should at least make sure the transformer is "good" and splice longer leads to the main connections inside the bell housing.

Next step: spreadsheet list of components and finding/ordering parts.

1PM-3:45PM, 8PM-11:30PM (6.25HRS)

LT Spice Fender Preamp Section Simulation

Today/tonight I drew up the preamp section for the Fender tone control in LT Spice. I added the appropriate models as SPICE directives (.inc) and pointed to them in LT Spice. I also took the opportunity to try out the potentiometer models. At the moment I have not plotted multiple variations of potentiometer values for the tone controls, however I plan to do so soon.

As the schematics show, I decided to go with the Super Reverb schematic for the preamp stage, "Vibrato" channel, with "Mid" tone control. This was a matter of taste, as I have always liked the overall voicing of the Blackface era Fender amplifiers.

I have also included a link to a free PC program that plots frequency response simulations for various tone control types. It is called Tone Stack Calculator. The program allows the user to substitute different values of components and get a visual representation of the overall frequency changes.

I plotted a graph for the component values used in my planned tube preamp stage. It looks quite close to the LT Spice simulation results.

Next steps: Plot multiple frequency responses for changes in potentiometer variation. Draw up the rest of the tube amp schematic, with component designators, and make a list of parts needed so I can begin the physical build.

4-7PM, 11PM-12AM (4HRS)

Worked on SPICE simulation today/tonight. I found a symbol and simulation model for a log potentiometer (audio taper) online, and attempted to teach myself to implement it in LTSPICE.

Why?

Well, the project I'm putting together uses log-taper pots, as do many audio-related electronic devices. They do so because the human ear hears changes in sound levels as a logarithmic function of voltage.

Because this is an electronic project for a college level course, I figure I may try to do some college level work. Hence the simulation.

LTSPICE was not taking my ".inc" SPICE directive and pointing to the model, so I copied the .SUBCKT (sub circuit) model data into a new file of its own and pointed the schematic symbol to that data instead. It worked!

I have a 100k-ohm audio taper stepped attenuator that I measured for a reference of resistance values. The model is practical enough for simulation use.

For use as a volume control, the simulation model does not give any valuable data, however, used with equalizer circuits, which this project contains, it would be interesting to plot values of the frequency changes as a result of resistance change.

LTSPICE has a function that allows a parameter of a device to be tested at varying values. I have drawn up a simple voltage-divider network using a logarithmic potentiometer model and ploted the DC voltage levels as a function of rotation. Note that I haven't figured out how to define a time step sequence of value change events. All the DC values are displayed on the same graph, but you can see that the levels are indeed plotted in a non-linear pattern. Had a linear pot been used, the DC voltage variations between one another would have been exactly the same.

 

Here's the model data for the log pot that I found. Credit shown to the originator.

* A Collection of Potentiometers
* ==============================
* Helmut Sennewald,                          12/23/2003         V1.1
*
*       pot_plog         exp(k*(1-x))
*
*       1 ____    1.0=wiper  
*             | 
*            | |  3
*            | |<---- wiper 0..1
*            | |
*       Rtap | |  Tap
*            | |
*       2 ____|   0.0=wiper
*
*
*     RTOT = total resistance
*     WIPER = ratio of travel of the wiper
*     RTAP = reference resistance at wiper=Tap
*            It is needed only for pot_plog, pot_nlog and pot_pow.
*            RTAP is measured between pin-2 and wiper.
*     TAP = ratio of travel when Rtap is reached


*---------- The Ideal Positive Logarithm Potentiometer --------------
*
* 1.0 <----- 0.0
*        |3
*      __V__
*  1--|_____|--2
*       
*  o--R1-o-R2--o
*
*  RTAP is resistance at travel TAP
*  Example: Rtot=10k, R=1 @ 0.001 
*           RTAP=1, TAP=0.001
*  RTAP and TAP define a point of the curve resistance versus ratio.
*
.SUBCKT pot_plog 1 2 3
* Parameters: Rtot, wiper, Rtap, tap
.param w=limit(0.01m,wiper,0.99999)
*
.param pwrexp=ln(Rtap/Rtot)/(1-tap)
.param ratio=exp(pwrexp*(1-w))
*
R1 1 3 {Rtot*(1-ratio)}
R2 3 2 {Rtot*(ratio)}
.ENDS

3:30-5:30; 11:00-1:30 (4.5HRS)

Worked on PCB layout for solid-state preamp stage. Need to make PCB:

- Print traces-only 1:1 layout on 2 laser printer transparency sheets

- Make 2-layer overlay of PCB pattern for pre-sensitized PCB material

- Expose PCB to UV light

- Develop PCB

- Cut to size, drill, clean

- Tin-plate PCB

10PM-12:30AM (2.5HRS)